You are on page 1of 224



Andrew S. Rivkin

Greenwood Guides to the Universe

Timothy F. Slater and Lauren V. Jones, Series Editors

An Imprint of ABC-CLIO, LLC
Copyright 2009 by Andrew S. Rivkin

All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise,
except for the inclusion of brief quotations in a review, without prior
permission in writing from the publisher.
Library of Congress Cataloging-in-Publication Data
Rivkin, Andrew S.
Asteroids, comets, and dwarf planets / Andrew S. Rivkin.
p. cm. (Greenwood guides to the universe)
Includes bibliographical references and index.
ISBN 978-0-313-34432-9 (hard copy: alk. paper)
ISBN 978-0-313-34433-6 (ebook)
1. Solar system. 2. Asteroids. 3. Comets. 4. Dwarf planets. I. Title.
QB501.R58 2009
523.2dc22 2009016114
ISBN: 978-0-313-34432-9
EISBN: 978-0-313-34433-6
13 12 11 10 9 1 2 3 4 5
This book is also available on the World Wide Web as an eBook.
Visit for details.
Greenwood Press
An Imprint of ABC-CLIO, LLC
130 Cremona Drive, P.O. Box 1911
Santa Barbara, California 93116-1911

This book is printed on acid-free paper

Manufactured in the United States of America

For my father

Series Foreword ix
Preface xi
Acknowledgments xiii
Introduction xv
1 A Matter of Definition 1
2 Historical Background 13
3 The Orbits and Dynamics of Small Bodies 27
4 Meteors, Meteorites, and Meteoroids 43
5 The Formation of the Solar System and the Small Bodies 57
6 Sizes, Shapes, and Companions of Small Bodies 69
7 Composition of Small Bodies 83
8 Surface Processes 101
9 Small Body Interiors 117
10 Small Body Atmospheres 131
11 Small Bodies and Hazards 141
12 Spacecraft Missions 155
13 Interrelations 171
Glossary 183
Annotated Bibliography 195
Index 203

Series Foreword

Not since the 1960s and the Apollo space program has the subject of astron-
omy so readily captured our interest and imagination. In just the past few
decades, a constellation of space telescopes, including the Hubble Space Tele-
scope, has peered deep into the farthest reaches of the universe and discov-
ered supermassive black holes residing in the centers of galaxies. Giant
telescopes on Earths highest mountaintops have spied planet-like objects
larger than Pluto lurking at the very edges of our solar system and have care-
fully measured the expansion rate of our universe. Meteorites with bacteria-
like fossil structures have spurred repeated missions to Mars with the
ultimate goal of sending humans to the red planet. Astronomers have recently
discovered hundreds more planets beyond our solar system. Such discoveries
give us a reason for capturing what we now understand about the cosmos in
these volumes, even as we prepare to peer deeper into the universes secrets.
As a discipline, astronomy covers a range of topics, stretching from the
central core of our own planet outward past the Sun and nearby stars to the
most distant galaxies of our universe. As such, this set of volumes systemati-
cally covers all the major structures and unifying themes of our evolving
universe. Each volume consists of a narrative discussion highlighting the
most important ideas about major celestial objects and how astronomers
have come to understand their nature and evolution. In addition to describ-
ing astronomers most current investigations, many volumes include per-
spectives on the historical and premodern understandings that have
motivated us to pursue deeper knowledge.
The ideas presented in these assembled volumes have been meticulously
researched and carefully written by experts to provide readers with the most
scientifically accurate information that is currently available. There are
some astronomical phenomena that we just do not understand very well,
and the authors have tried to distinguish between which theories have wide
consensus and which are still as yet unconfirmed. Because astronomy is
a rapidly advancing science, it is almost certain that some of the concepts
presented in these pages will become obsolete as advances in technology
yield previously unknown information. Astronomers share and value a


worldview in which our knowledge is subject to change as the scientific

enterprise makes new and better observations of our universe. Our under-
standing of the cosmos evolves over time, just as the universe evolves, and
what we learn tomorrow depends on the insightful efforts of dedicated sci-
entists from yesterday and today. We hope that these volumes reflect the
deep respect we have for the scholars who have worked, are working, and
will work diligently in the public service to uncover the secrets of the
Lauren V. Jones, Ph.D.
Timothy F. Slater, Ph.D.
University of Wyoming
Series Editors

A great deal of attention has of late been directed to the smaller members of
the solar system. In the past decade, we have seen spacecraft orbit and land
on asteroids, and smash into comets. We have taken our first steps to better
understand and characterize the threat to our civilization posed by collisions
with near-Earth objects. For the first time in a generation, the sky was graced
with extraordinarily bright comets. And scientists have begun to discover
and catalog new planet-size objects in the far reaches of the outer solar sys-
tem, leading to controversy about what exactly makes something a planet.
These new findings have been met with interest and enthusiasm by the
general public. The prevention of and possible consequences of asteroid
impact have been the central plot lines of popular movies, and the contro-
versy over the status of Pluto has been immortalized in Web site petitions
and T-shirts.
This increase in attention comes hand in hand with an increase in our
knowledge about the small bodies of the solar system. The combination of
new telescopes, more capable computer simulations, sophisticated labora-
tory techniques, and spacecraft data has made this a golden age for asteroid
and comet studies. We have long known that these objects have much to tell
us about the solar system, its history, and its evolution. We are only now
able to understand large swaths of the story being told, and there is much
more yet to be understood.
A volume in the Greenwood Guides to the Universe Series, Asteroids,
Comets, and Dwarf Planets is intended as an introduction to the current
state of knowledge; the volume is aimed at a nontechnical audience of stu-
dents who do not plan to specialize in astronomy, geology, or physics, and
at public library patrons who are interested in the field and seeking current
information. The volume covers a wide range of topics of interest for these
readers, if only at a level to whet a readers appetite rather than provide in-
depth knowledge. In contrast to a separate treatment of each population,
the focus of this book is on processes and features and a comparison and
contrast of how comets, asteroids, and dwarf planets are affected by those
processes and to what extent they share those features. Also explained are


the techniques and logic that have led planetary scientists to the conclusions
that they have reached.
Containing 13 chapters and an Introduction to the topic, Asteroids, Com-
ets, and Dwarf Planets begins by looking at the factors that set those objects
off from the major planets. After a tour of the history of their investigation,
we look at where the small bodies are found in the solar system, and what
we have learned from the samples in our laboratories about their nature
and formation. Following this come several chapters that look at the com-
position and structure of asteroids, comets, and dwarf planets, from their
interiors to their surfaces, the tenuous atmospheres suspected on the largest
objects, and the satellites that accompany some objects. Finally, we turn to
the hazard posed by some of the small bodies and the different types of
space missions that have been sent to them and what we have learned,
before considering the interrelations between small bodies and other popu-
lations both inside and outside the solar system.
The chapters contain sidebars that offer more in-depth information on a
broad array of important and engaging topics. The sidebars are included to
give readers a deeper understanding of various issues and questions without
interrupting the main flow of the chapter narrative. The chapters also con-
tain numerous photographs and other illustrations that serve to augment
the discussion in the text. Terms highlighted in boldface in the text alert the
reader to the terms inclusion in a Glossary, which offers a brief, useful
description of the term as currently used in the field. Each chapter con-
cludes with a brief list of further print and electronic information sources
for that topic; a bibliography at the end of the book lists more general infor-
mation sources. The volume concludes with a detailed subject index.

I would like to acknowledge the help and support of a number of people

without whom writing this book would have been more difficult and whose
input was a big help.
Initial thanks go to the team at Greenwood Press (now ABC-CLIO)
Lauren Jones, Kevin Downing, and John Wagner. I greatly appreciate their
patience, comments, and encouragement. I would also like to thank Tim
Spahr, without whom I would not have had the opportunity to participate
in this series.
My colleague Ralph Lorenz was a great resource in many stages of this
process. His extensive writing experience greatly eased my burden by assist-
ing me in identifying possible problems and their solutions well before I
would have spotted them. My colleagues Neil Dello Russo and Nancy Cha-
bot also contributed by lending research materials and answering my ques-
tions about the many specialty areas included in the book with which I have
less personal experience.
Thanks also to my employers at the Johns Hopkins University Applied
Physics Laboratory, particularly Andy Cheng, Ben Bussey, and Louise
Prockter, for their support and encouragement.


Current scientific research about asteroids, comets, and dwarf planets takes
many forms. While one might imagine astronomers spending lonely nights
at observatories on remote mountaintops, many of the latest advances have
been made by someone using a laptop at their desk in a major city. It is the
constant interplay between theory, modeling, observations, and experi-
ments that makes planetary science in general, and small bodies studies in
particular, so dynamic. It is important to note that despite the division of
research into the following separate areas, many projects take advantage of
more than one type of research and many scientists participate, for instance,
in both observing and modeling. The bulk of Asteroids, Comets, and Dwarf
Planets will focus on the current state of our knowledge of the small bodies,
so we take a moment here to consider the general ways this scientific
research is performed.
The first studies of the small bodies were done by observers, using their
eyes alone, whose names are lost to history. In the intervening centuries,
increasingly sophisticated instruments have been used in the discovery and
study of comets, asteroids, and dwarf planets, first from the Earth, and
eventually from space and even asteroid surfaces. At present, useful scien-
tific observations are being made from several spacecraft and telescopes
around the world, ranging from the largest that have been built to small
homemade versions operated by hobbyists.


The most visible and sophisticated means of learning about comets, aster-
oids, and dwarf planets is via space missions. Nearly a dozen small bodies
have been visited by spacecraft, starting with a fleet of missions visiting
Comet Halley in the 1980s. More missions are en route, including studies
of the dwarf planets Ceres and Pluto in the coming decade.
These missions have provided surface imagery in great detail, which has
been a boon to geologists and geophysicists seeking to understand the his-
tories and processes on small bodies. They have also returned samples of


cometary dust, and missions have been proposed to bring back surface sam-
ples of asteroids and comets. The quality of spacecraft data has revolution-
ized our knowledge of the small bodies.


Unfortunately, space missions are complex, expensive undertakings.

Because of this expense and difficulty, only a handful of comets and aster-
oids have been visited by spacecraft, and the vast majority will never be
explored in that way. Therefore, it is still the case that telescopes are used to
collect most of the information we have for comets, asteroids, and dwarf
planets. Well over 100,000 objects are known and catalogued, with several
hundred thousand more having less well-known orbits. Most of these objects
were discovered by Earth-based telescopes with diameters of 11.5 m. For
comparison, the Hubble Space Telescope is 2.4 m in diameter, and the largest
telescopes on Earth have sizes up to 10 m. Since the late 1800s, discoveries
have been made using photographic techniques rather than by using an
eyepiece, originally using film and moving to digital cameras in the 1980s.
Studies of the physical properties of small bodies can require larger tele-
scopes, although there are many astronomy hobbyists who help advance
science by making simple, but much-needed, observations on personally
owned telescopes in their spare time. However, more difficult observations
of fainter, smaller, and more distant objects use the largest facilities available
to astronomers. Two of the more prominent facilities include the telescopes
at Keck Observatory in Hawaii and the radar facilities at Arecibo Observa-
tory in Puerto Rico.


While telescopic data is a critical component of the study of asteroids, com-

ets, and dwarf planets, scientists have also been able to perform laboratory
analyses and experiments to gain further insight. Meteorites, pieces of aster-
oids that fall to earth, are often the subjects of these analyses, as the most
modern and sophisticated equipment available to geochemists determines
their elemental compositions and identifies particular textures and minerals
that give clues to the history of the meteorite and the entire solar system.
Experiments using sophisticated furnaces and chambers simulate the tem-
peratures, pressures, and compositions present when the meteorites
formed. Experiments are also performed to study the effects of the solar
wind and micrometeorite bombardment of asteroid and transneptunian
objects on their apparent compositions. Some scientists perform experi-
ments of a more energetic nature. These researchers seek to learn about
impact processes in the solar system, and use guns designed to fire pellets at
Introduction  xvii

speeds of over 1 km/s (or over 3,600 km/hr). The lessons learned about the
breakup of small objects are applied to the cratered and disrupted popula-
tion of small bodies. In a sense, the Deep Impact mission was just a large-
scale collisional experiment!


Even small bodies are too large to allow experiments on the proper scale.
For instance, the largest human-made craters on Earth have been made
during nuclear tests and are still exceedingly small by astronomical stand-
ards. The effects of large impacts must be studied through mathematical
extrapolation based on small craters and a knowledge of physics. The same
sort of logic applies to studies of long-term processes that can take thou-
sands to millions (or billions!) of years to complete.
The increased capabilities and availability of computing power has
helped scientific research immensely. The small bodies community has
turned this computing power to simulations of the gravitational interac-
tions of large populations, leading to new theories about solar system for-
mation, the migration of asteroids and comets through the solar system,
the threat posed by the impact of near-Earth objects, and the formation and
evolution of binary asteroids, to name a few applications.


Finally, there are those whose research requires only a word processor or a
whiteboard. The theoretical underpinnings of small bodies research are im-
portant areas and have shown particular advancement in recent years. The
recognition that nongravitational forces like the Yarkovsky and YORP forces
are important has flown from new theoretical appreciation. Consideration
of early solar system history has led to new expectations for the properties
and history of asteroids, comets, and dwarf planets. These expectations feed
back into new proposed observations, experiments, and modeling efforts.
The following pages compile the results of decades of research, using all
the techniques described herein. Science is always changing, in particular a
field like small bodies studies, where a new meteorite fall or a bright comet or
data returned from a spacecraft can be drastically different from previous ex-
perience. The aim of this volume is to provide a general guide to our current
understanding, upon which new findings can be added and understood.
A Matter of Definition

According to the International Astronomical Union (IAU), the solar system

contains one star, eight planets, at least five dwarf planets, and uncounted
numbers of small solar system bodies. According to others, however, there
are nine planets in the solar system. Still others count at least 10 planets, if
not more. In this chapter, we will attempt to define a planet, or at least
show how one might do so. We will also follow the evolution of this debate,
and consider the pros and cons for each camp. We will then turn to the
small bodies to lay a foundation for further discussions throughout this
How might we go about defining a planet? An admittedly circular way
might be to use the objects we already call planets and try and figure out
what they have in common. We might use characteristics like their orbits,
sizes, and shapes.
For the ancient Babylonians and Greeks, little information was available,
but one thing was clearof the objects in the sky, most remained in place
with respect to one another, while a small number moved. These moving
objects, which we now call Mercury, Venus, Mars, Jupiter, and Saturn, as
well as the Sun and Moon, were all considered wandering stars, or planetes
asteres in Greek. Comets were known to the ancients, but considered either
supernatural or related to weather in some unknown way. As the heliocen-
tric model gained acceptance and the Sun replaced the Earth at the center
of the solar system, the two bodies swapped classifications, with the Earth
being regarded as a planet, while the Sun was not. The word planet was fur-
ther refined through the seventeenth and eighteenth centuries as objects
were found orbiting Jupiter and Saturn; these objects were first called


planets, but the term satellite eventually came into use for anything orbiting
something other than the Sun, including our Moon. The discovery of
Uranus in 1781 added a new planet to the scientific consciousness for the
first time.
As described in more detail in Chapter 2, the first four of what we now
call asteroids were discovered in the early 1800s. Each discovery was
treated as a new full-fledged planet, and individual symbols were assigned
to each body, just as had been done for the other planets. Observations
quickly determined that this quartet was not like the rest of the planets,
however. The first curiosity was the similarity of their orbits, which were
much closer to each other than any other sets of planets. Furthermore,
while most planets showed disks when viewed through telescopes, these
four objects remained stubbornly starlike in appearance regardless of
magnification. This fact led William Herschel to dub them asteroids,
meaning starlike.
By the 1850s, with over a dozen asteroids discovered, the general consen-
sus was that the asteroids were fundamentally different from other planets,
and they were considered their own class: minor planets. The major plan-
ets, by comparison, were large and orbited by themselves, along with any
Thus when Pluto was discovered in 1930, there was no question in any-
ones mind that it was a planet. It was believed to have a significant mass
and did not share its orbit with any other objects (even though it crossed
Neptunes orbit, which attracted some notice). However, as we learned
more about the planets and as estimates of Plutos mass got smaller (and
predictions for objects in the transneptunian region became stronger),
astronomers began to question whether Pluto was really a planet.


When learning biology, students are presented with a few simple rules to
determine whether something is a mammal: Does it give birth to live
young rather than lay eggs? Does it feed its young milk? Does it have
hair? Similarly, one could look at those objects that are undeniably plan-
ets and try and come up with a similar set of questions. However, this is
not as straightforward as it may seem, as we can see from the following

 Does it have an atmosphere? The giant planets, Venus, Earth, and Mars, all
have atmospheres. It is arguable whether Mercurys thin sodium and potas-
sium atmosphere should count in this way, though if it is considered to have
one, so do a great many other objects including the Moon (as is further dis-
cussed in Chapter 10). There is no question that Saturns satellite Titan has an
atmosphere, however, although it is not generally considered to be a planet.
A Matter of Definition  3

 Does it orbit the Sun rather than another planet? Many objects not normally
considered to be planets orbit the Sun, such as asteroids and comets. This
suggested rule would allow us to rule out Titan as a planet, but would still
need to be combined with other rules.
 Does it have any satellites of its own? Again, there are objects known to have
satellites that are not considered planets, for instance, the asteroid Ida (among
others, as discussed in Chapter 6). Conversely, neither Mercury nor Venus
have any satellites, and there is general consensus that they are planets.

Obviously, to try and define a planet in this manner will require several differ-
ent questions to be answered, and agreement on which are the important ques-
tions and which are less useful. Perhaps there is another way to define a planet.


Interestingly, the question, What is a star? does have a simple answerA

star is an object generating energy through nuclear fusion. Astronomers
have been able to make observations of bodies and measure their tempera-
tures, which, in turn, can easily let us calculate whether an object is a star.
Ideally, there would be a similar measurement or set of measurements
that can be made to determine whether an object is a planet or not. For a
long time, the mass of an object was informally treated as the discriminator
between objects that are planets and those that are not. Above a certain size,
about 13 times more massive than Jupiter, objects will undergo fusion and
become stars. Planets, then, are 13 Jupiter masses or smaller. But what
about the small end?
When looking at the sizes and masses of solar system objects, there is one
place in the distribution where a break between planets and non-planets
might naturally be placed since it is free of objects and it is unlikely that any
objects will be found in the gap in the futureUranus is 14 times more
massive than the Earth and 4 times larger in radius. Some have argued that
the classification that makes the most sense is for the solar system to have
four planets: Jupiter, Saturn, Uranus, and Neptune, and everything else,
including the Earth, is unworthy of the name! Few have ever taken this sug-
gestion seriously, although that may be because as earthlings we are biased
against it. Figure 1.1 shows the relative sizes of the planets (including
Pluto); where might we draw a line?
The size of Pluto was long informally treated as the low-mass end for
planets (again, with the usual restriction that larger-mass satellites like
Ganymede and Titan were ineligible for planethood). The first reliable mea-
surement of Plutos mass came in the late 1970s, after the discovery of its sat-
ellite Charon. Plutos mass was found to be by far the smallest of any planet,
including Mercury. Through the 1990s and early 2000s, a group of objects
orbiting beyond Neptune and called transneptunian objects (TNOs) were

The nine bodies generally recognized as planets in the late twentieth century
Figure 1.1.
have a wide range of sizes, shown here to scale (although their distances from one
another and the Sun are not to scale). Pluto, at the far right, is much smaller than any
of the others, and is much more similar in size to large transneptunian objects discov-
ered in the past 10 years. The difference in size between the jovian planets (Jupiter, Sat-
urn, Uranus, and Neptune) and the terrestrial planets (Mercury, Venus, Earth, and
Mars) is as striking as the difference between Pluto and Mercury, the smallest terrestrial
planet. NASA/Lunar and Planetary Laboratory.

found with sizes (and estimated masses) ever closer to Plutos, until finally
the discovery of Eris showed that objects larger than Pluto were still to be
found. Was Eris the tenth planet? Or did this show that Pluto was simply the
first TNO to be discovered and not worthy of a planetary classification?
The answer depends in a lot of ways on what you think a planet is, which is
the question we seek to answer in the first place.

Whats in a Name?
The naming of various planets and dwarf planets has caused a surprising amount of controversy
and discord over the years. In the seventeenth century, Galileo proposed to honor his patrons by
naming what are now called the Galilean satellites of Jupiter the Medicean planets after the Med-
ici family. Similarly, in the eighteenth century, William Herschel called his discovery Georgium
Sidus (The Georgian Star) after King George III of England. The French, understandably reluctant
to name a new planet after the king of England, preferred the name Herschel, with the German
preference of Uranus eventually winning out. Ceres was originally called Ceres Ferdinandea, in
partial homage to King Ferdinand III of Sicily. Here too, the reference to an earthly ruler was omit-
ted. Finally, the discovery of Pluto led to an outpouring of suggested names, including Minerva,
Cronus, and the somewhat unlikely Constance (proposed by Constance Lowell, widow of Percival
Lowell, astronomer and founder of the Lowell Observatory, where Pluto was discovered).
Naming controversies have continued to this day among the asteroids. Particular uproar faced
the naming of 2309 Mr. Spock, named not after the Star Trek character, but after the discoverers
cat. This led to a ban on asteroids named after animal pets. The naming of Eris (goddess of dis-
cord) and its satellite Dysnomia (lawlessness) was a nod toward the classification controversy they
helped bring to a head.
A Matter of Definition  5


The potentially circular nature of defining planethood is considered an

unsolvable problem by many astronomers. They dont feel that a definition
is necessary, arguing that any distinction between planets and non-planets
is arbitrary and not important scientifically. The word planet already has a
well-established meaning in many languages, and while it may not be a pre-
cise analytical meaning, it is not, astronomers argue, their place to redefine
this word.
An example of this viewpoint is to think of the definition of a book. If
asked to describe a book, many people will come up with similar descrip-
tions. However, the details may disagree. Some will argue that a book must
have content of some kind, so blank books and phone books are not books.
Others would require books to be physical objects, so an electronic book
would not count. Nevertheless, most people could identify most things they
encountered as either a book or not a book. Similarly, some would argue,
the details of what is a planet dont matter, because scientists basically agree
whether a specific object is a planet even if they dont all agree exactly on
the details of the classification. They see further parallels in the definitions
of life or art (or obscenityas a U.S. Supreme Court justice once noted, he
wouldnt define obscenity, but he knew it when he saw it), which at their
extremes have no consensus.
On the other hand, supporters of some kind of definition note that many
words in common use also have specific scientific definitions, such as sand.
They argue that to reduce confusion in scientific work, a definition is


For many astronomers, the importance of a definition is that it should

quickly allow a determination to be made for an object, with no ambiguity
(at least in principle). This suggests a reliance on observable properties such
as brightness, mass, size, or orbit.
The shape of an object has been used as a basis for classification. The
largest objects in the solar system are basically spherical objects. Why is this
so? We have all heard of sea level. The oceans of the Earth have surfaces at
the same relative height. Water in rivers and streams on the Earths surface
flows downhill until it reaches the ocean and sea level. If somehow the sur-
face of the Atlantic Ocean were higher than the Pacific and Indian Oceans,
water would flow downhill out of the Atlantic and into the other oceans
until they were once again at the same level. If water were added to the
oceans, via melting ice or other means, the water would distribute itself
around the world until a new level was reached, higher in comparison to
the land but still equal across all the oceans.

Imagine a planet with no dry land at all, only water. Everything at its sur-
face would be at the same levelsea level. The distance from any point on
its surface to the center of the planet would be the same. In other words, the
planet would be spherical. A planet does not have to be covered in water for
this to be true, however. A fluid planet with a thick atmosphere and no real
solid surface, like Jupiter or Saturn, will also tend to be spherical for similar
Even solid bodies like the Earth will become spherical. The large amount
of mass in the Earth can act like a fluid over a course of years. As a result,
the difference in elevation from the highest point on the Earths surface
(Mt. Everest) to the lowest point under the ocean (in the Marianas Trench)
is less than 0.2 percent of the mean radius of the Earth. By these standards,
the Earth is very smooth indeed. Mars, with higher peaks and deeper val-
leys, is still smooth within 1 percent of its radius.
Planets are not exactly round, however. Precise observations of Jupiter, or
even the Earth, show that their exact shapes are oblate, with the distance
from north pole to south pole slightly less than the equatorial diameter.
This is because planets rotate. For instance, all objects on the Earth have an
angular speed of 360 degrees per 24 hours (in other words, they complete
one rotation in a day). The linear speed can vary dramatically, however;
objects on the equator travel on a circle with a circumference of roughly
40,000 km every day with respect to the center of the Earth, nearly
1,700 km/h. Objects at the north or south poles, however, dont move at all
with respect to the center of the Earth. Objects at various points between
the Equator and poles move at intermediate speeds depending on their lati-
tudes. The extra speed felt by objects on a planetary surface due to rotation
causes an extra force, which makes material flow toward the equator, caus-
ing planets to be not-quite spherical. Objects that act in this way are said to
be in hydrostatic equilibrium. Large masses like the planets, and smaller
fluid masses, are found to be in hydrostatic equilibrium. Smaller solid
masses, however, cannot overcome their own strength and can retain irreg-
ular shapes (like the asteroids Ida or Eros, among many others).
Whether an object is massive enough to be in hydrostatic equilibrium is
seen by many astronomers as a key criterion for planethood. For massive
enough objectslike Jupiter, Neptune, or Venushydrostatic equilibrium
is guaranteed. At the small end, observations might be necessary to deter-
mine whether an object is massive enough (or is the right shape, if mass
measurements are not available) to be in hydrostatic equilibrium.
A proposal was made in 2006 to the International Astronomical Union,
which would have defined a planet as an object that had the following

1. Sufficient mass to be in hydrostatic equilibrium

2. Orbited a star
3. Is neither a star itself nor a planetary satellite
A Matter of Definition  7

Figure 1.2. Two non-planets appear in this figure. At left is Saturns satellite Titan, which is larger
than the planet Mercury and has a thick atmosphere. It is large enough to be in hydrostatic equilib-
rium, one of the criteria for planethood in the recent IAU definition, but because it is in orbit
around Saturn, it is not eligible. At right is the asteroid Eros, visited by the NEAR Shoemaker space-
craft. It is much smaller than Titan, too small for its gravity to have pulled it into a spherical shape.
It is also much too small for its gravity to affect other objects in nearby orbits. It therefore fails both
tests for planethood. However, classification as a non-planet does not diminish the scientific value
and interest in these or the other large satellites and small bodies of the solar system. Left, NASA/
JPL/Space Science Institute; right, NASA/JPL/JHUAPL.

In addition, a category of double planet was proposed, if two compo-

nents of a system were each in hydrostatic equilibrium, had similar
masses, and a particular orbital configuration. Pluto and Charon would
have qualified as a double planet in this definition. In addition, Ceres and
Eris would have qualified as planets, giving the solar system 12 known
planets. A dozen or more additional objects, almost all beyond Neptune,
might also have been classified as planets, pending the collection of addi-
tional data.
This proposed definition did not satisfy everyone. Major and minor
problems were pointed out, with some people uncomfortable with the idea
that the solar system might have dozens of planets. Others argued that stud-
ies showed Pluto to be a typical TNO and convenience and history were no
reason to confer special status, regardless of its shape.


What would be a reason to confer special status? One proposed classifica-

tion looks to the formation of the solar system, discussed in further detail

in Chapter 5. The eight undisputed planets from Mercury to Neptune all

were very effective in either collecting (or accreting) mass, capturing it into
satellite orbits, or ejecting it to other parts of the solar system (or out of the
solar system entirely). By the end of solar system formation, the vast major-
ity of material near the Earths orbit was part of the Earth itself. The same is
true for the other planets, from Mercury to Neptune. This process, infor-
mally dubbed clearing of neighborhoods, is considered by some a critical
test of planethood.
A quantitative measure of the ability of an object to clear its neighbor-
hood is provided by a parameter called L (the Greek capital letter lambda).
This is equal to an objects mass (M) squared times a constant (k) divided
by its orbital period (P): L = kM2/P. We can compare different objects by
measuring masses in Earth masses and the period in years, setting k = 1.
Then L = 1 for the Earth. Jupiter, over 300 times the mass of the Earth and
with a period of nearly 12 years, has a value of L roughly equal to (300)2/12,
or 7,500. Neptunes mass is 17 times that of the Earth, its orbital period is
just short of 165 years, and its L is roughly 1.8. For Mars, L = 0.006, appa-
rently quite small but still above the threshold expected for neighborhood
Pluto, which orbits relatively close to other objects of similar mass,
fails this test. Its mass is only two thousandths that of the Earth, and its
period is nearly 250 years. As a result, L for Earth is 50 million times
larger than for Pluto, not nearly enough for Pluto to have any real dy-
namical effect on nearby objects. Dynamicists, who wish to study the
orbits of asteroids and comets over millions of years or longer, need to
include the gravitational effects of the major planets to have accurate
results, but find that they can ignore Pluto in their studies without any
loss of accuracy.
Using the criterion that an object must have cleared its neighborhood,
as well as the previous requirements for hydrostatic equilibrium and not or-
biting another planet, we are left with eight planets: Mercury, Venus, Earth,
Mars, Jupiter, Saturn, Uranus, and Neptune.
This dynamics-influenced definition has been adopted by the Interna-
tional Astronomical Union. It also created a category called dwarf planets,
which are objects that are large enough to be in hydrostatic equilibrium,
but are not capable of clearing their neighborhoods. This category origi-
nally included Pluto, Ceres, and Eris, although two additional objects have
been officially included into this category since the original announcement,
and more objects could be classified as dwarf planets pending additional
data. An additional category was added to the IAU classification scheme in
2008, with dwarf planets orbiting beyond Neptune being further classed as
plutoids. Thus Eris, for instance, is both a dwarf planet and a plutoid, while
Ceres is only a dwarf planet. Everything else in the solar system that orbits
the Sun rather than a planet is then classified as a Small Solar System Body
A Matter of Definition  9


While there is now an official IAU definition of a planet, it is not certain as of

this writing that the last word has been spoken. The most common concern
with the official definition is that it is vaguely worded, and based on models
and simulations in a critical way, and those models could change with time.
A definition based only on observations, like the ones based only on shape,
would not suffer from that drawback. In addition, it has been noted that
using the IAU definition and the models for neighborhood clearing, an object
the size of Mars found at the distance of Eris would not qualify as a planet,
and indeed a hypothetical Earth-sized object at a large enough distance from
the Sun would not be considered a planet using that criterion either.
An additional complaint is that by the technical wording of the IAU defi-
nition, technically speaking there are no planets outside the solar system. It
would be possible to change some of the language to allow planets around
other stars, but without knowing the details of those systems, it would be
difficult or impossible to determine whether any of those objects have
cleared their neighborhoods, and difficult or impossible to apply the IAU
definition. It is also hard to see how any free floating planets, planet-size
objects that may have escaped the solar systems in which they formed early
in their history, could be accommodated by a definition that includes dy-
namics. All told, it appears likely that the IAU definition of a planet will be
refined in the coming years, and it is possible that researchers will continue
to use definitions that do not match any official definition at all.


While the definitions of groups of small bodies are not as contentious as

those for planets, they are in some ways no less murky. Asteroids and com-
ets have been distinguished from one another for centuries by a simple
observational question: Does the object appear starlike or fuzzy through a
telescope? As noted before, asteroids were named for their starlike appear-
ance, and are point sources (that is, they are so small they appear as points
of light regardless of magnification) in all but a very few telescopes, includ-
ing the Hubble Space Telescope. Comets were named for their fuzzy
comae and tails, the word comet meaning hairy star in ancient Greek.
While the ancient Greeks only had their eyes for observing, this distinction
between asteroids and comets has been maintained to date, with any object
showing evidence of a coma or tail being classified as a comet. However,
some objects have been found to sometimes appear starlike and sometimes
have a coma. If an object appears cometary when discovered, it will main-
tain this classification. If, however, an asteroidal object is later seen to have
a cometary appearance, it can be given a cometary classification as well as
an asteroidal one. To date there are three such objects.

By the late-twentieth century, it was recognized that comets were largely

icy bodies that originated in the outer solar system, and asteroids were
largely rocky bodies that originated in the inner solar system (see Chapter 5).
This has led to a separate, informal distinction between comets and asteroids.
For some researchers, asteroid has become shorthand for a rocky object, and
comet for an icy one. Alternately, asteroids represent bodies that originated in
the inner solar system, and comets in the outer solar system.
Unfortunately, these two distinct ways of separating comets and asteroids
can lead to occasional confusion. One will sometimes hear of dead
comets, which show no evidence of a coma or tail (and thus would be aster-
oids by the observational definition), but are believed to have originated
in the outer solar system (and would thus be comets by the compositional/
dynamical definition). In the chapters that follow, we will use the composi-
tional/dynamical definition unless otherwise notedasteroids are generally
rocky and metallic objects believed to originate in the inner solar system;
comets are generally icy bodies believed to originate in the outer solar sys-
tem. In some cases, the line is blurry, particularly for some populations of
bodies for which we currently know little. We will address the relationship
between asteroids and comets, and the objects that straddle the lines
between them, in Chapter 13.
Finally, another term we will often use in later chapters is transneptunian
object. As mentioned previously and suggested by their name, TNOs are
found beyond the orbit of Neptune. These are thought to be similar in com-
position to comets in general, but are named to indicate that their current
orbits are far from the Sun.


The classification of the bodies of the solar system is a surprisingly subjec-

tive process. There are several different underpinnings for defining planets
and non-planets, each with strengths and drawbacks, and no objectively
correct answer. The diversity of the small body population is reflected in
the diversity of how we classify themcomets and asteroids are distin-
guished from one another by composition and dynamical history, as well as
by centuries-old observational descriptions. The new category of dwarf
planet is based on shape and theoretical models of gravitational influence.
TNOs are defined based on their orbits.


Scientific American article about the Planet Debate:

A Matter of Definition  11

A history of the status of Pluto:

Video of IAU debate:
Commentary on the IAU debate from a dissenting astronomer: http://www.lowell.
NASA page about the definition of planets and dwarf planets: http://solarsystem.
Historical Background

The small bodies of the solar system have been of interest to astronomers
for centuries. While we continue to learn more about them with each pass-
ing year, in this chapter we will focus on the earliest stages of their study.
We will consider how asteroids and comets were interpreted and how those
interpretations were shaped by then-current theories, and how they in turn
shaped how those theories evolved.


Astronomy is a modern science with ancient roots. Our ancestors were

familiar with the night sky. Living without electricity in relatively small
groups, and critically dependent upon understanding the seasons in order
to plant and reap their crops at the right times, they gained an intimate
knowledge of the cycles overhead. They could see thousands of stars, whose
rising and setting times slowly changed through the year, but whose posi-
tions relative to one another were unchanging.
They also knew of lights in the sky that moved relative to the stars.
Among them were the two brightest objects in the skythe Sun and the
Moon. In addition, there were five bright objects that appeared like stars to
their naked eyes but did not remain fixed. These wandering stars, called
planets in Greek, are known to this day by the names of Roman gods:
Mercury, Venus, Mars, Jupiter, and Saturn. The stars and these seven wan-
derers acted in predictable ways, and the ancients of many cultures tracked
them day after day, year after year, and century after century.


Occasionally, however, this predictability was unexpectedly disturbed.

The stargazers of the time, who believed studying the movement of the stars
and planets could lead to an ability to predict the future (a nonscientific
belief that has evolved into modern astrology), were often rattled by these
disturbances, and they connected floods, wars, epidemics, famines and the
likes to these bad stars or disasters in Latin.
The most common of these bad omens were the objects we know as com-
ets. Appearing without warning and looking like brooms, or fiery swords,
comets were often noted in conjunction with the deaths of kings, including
Genghis Khan and Julius Caesar. A comet seen in 66 AD was believed to be
an omen of Jerusalems destruction, which occurred only a few years later.
Perhaps the most famous association of a comet with the fates of kingdoms
was the appearance of Comet Halley in 1066. Harold, King of England, saw
the comet as a bad omen, while it raised the morale of William of Nor-
mandy and his successful invading troops. The events of the Norman inva-
sion of England were recorded in the Bayeux Tapestry, which includes
depictions of the comet.


As the modern era of Western science began, observations of comets began

to be used to learn quantitative facts about our universe rather than seen as
portents of good or ill fortune. After the sixteenth century, when Coperni-
cus suggested that the Earth orbited the Sun, rather than vice versa, other
astronomers continued the work of trying to figure out how the stars and
planets actually moved. Without a conception of gravity, it was thought
since the time of the ancient Greeks that the stars and planets all were physi-
cally attached to transparent spheres (often called crystalline spheres),
with one sphere per planet and an additional one for the stars.
The existence and behavior of comets directly challenged the idea of crys-
talline spheresnot only would each comet require its own sphere, but
comets seemed to move over a variety of distances, making it seem that they
would smash into the other planetary spheres during their visits. For this
reason, many astronomers concluded that comets were not astronomical
objects, but were simply an unusual type of long-lived weather, generated
in the Earths atmosphere and just appearing to be among the planets. This
view was popularized by Aristotle around 350 BC (who also first named
these objects kometes, meaning hairy star in ancient Greek, after their
In the late sixteenth century, Tycho Brahe, the greatest astronomer of his
era, performed painstaking observations of the sky. The precision of his
measurements, made in the last decades before telescopes were invented,
allowed him to measure the parallax, or apparent angular shift, of celestial
bodies. An easy illustration of parallax is as follows: Close one eye and hold
Historical Background  15

your thumb so it is blocking an object across the room from you. Now
switch the eyes that are open and shut, and notice that your thumb is no
longer blocking the object. The apparent motion of your thumb is greater if
it is held closer to your face.
This same principle can be used for objects in the sky. The Moon, for
instance, will be seen in a slightly different position relative to the stars
when viewed on opposite sides of the Earth. In Tychos time, collaboration
with Chinese or Islamic astronomers was not possible, but because of the
Earths rotation, Tycho could make all of the necessary observations himself
and remove the effect of the Moons rotation to obtain the correct answer.
A comet that appeared in 1577 provided another object for which Tycho
could measure parallax (see Figure 2.1). However, he found no measurable
parallax, implying a distance beyond the Moon and certainly outside of the
atmosphere. These observations and conclusions showed that the crystal-
line spheres could not exist.
Tycho died before telescopes were used for astronomy, and the Comet
of 1618 was the first to be observed with the new tool. Galileo Galilei,

Figure 2.1This woodcut illustrates the Great Comet of 1577, shown by Tycho Brahe to be
a solar system body rather than a weather phenomenon, as comets were thought to be
by scholars of the time. This comet is now more than 10 times further from the Sun than
Neptune is. Joerg P. Anders/Bildarchiv Preussischer Kulturbesitz/Art Resource, New York.

perhaps the most famous astronomer of all time, made observations.

However, the first telescopic observations are usually credited to Johann
Baptist Cysat, a Swiss Jesuit astronomer. Cysat was the first to distinguish
the nucleus of a comet, and demonstrated that the comets orbit was
more parabolic than circular, indicating that its aphelion was exceedingly


By the end of the seventeenth century, great advances in astronomical

theory by Johannes Kepler and Isaac Newton established the Universal
Theory of Gravitation, and eliminated the need for any crystalline spheres.
Newtons friend Edmund Halley noticed a pattern when looking at come-
tary appearances, and a similarity between comets that appeared in 1456,
1531, 1607, and 1682. Halley hypothesized that comets could be in orbits
that made relatively frequent returns, and that all of those comets were in
fact one and the same. He predicted this comet would return in 1758, a cor-
rect prediction that would lead to the comet taking his name: Halleys
Comet, known as Comet Halley to astronomers.
Around this same time, telescopes began to be used in making new
comet discoveries instead of merely making observations of known
objects. Dedicated searches of the sky had been performed for centuries
by Chinese astronomers, but the invention of the telescope now enabled
Europeans to make discoveries much more frequently. The astronomical
pastime of comet hunting, popular among hobbyists to this day, had a
number of immediate effects on both astronomical culture and astronom-
ical paradigms.
In the first case, a number of astronomers found that some objects in the
sky looked like comets but were fixed among the stars. One particularly
active comet hunter, Charles Messier, compiled and published a list of
objects that were prone to be confused with comets in the relatively small
telescopes available in the 1770s. This list, now known as the Messier Cata-
log, was the first collection of what were simply called nebulae (or
clouds), but are now known to include galaxies, globular clusters, and plan-
etary and star-forming nebulae. Thus, comet hunting led to the first study
of these deep-sky objects.
Comet hunting also cemented a realization that there were potentially
many objects in the universe that were yet undiscovered. In 1781, William
Herschel found what he thought was a comet, although further observa-
tions revealed an exceedingly distant orbit, no tail, and a disk. Rather than a
comet, Herschel had discovered the planet Uranusthe first planet to be
discovered in human history. There is evidence that Galileo observed the
planet Neptune when it passed close to Jupiter in 1612, and he even noted
that it moved, but since there was no expectation that any unfound planets
Historical Background  17

existed (and indeed, Galileo still had lingering doubts about Tychos con-
clusions about comets), Galileo did not realize what he had found. By
Herschels time, astronomers accepted the potential for yet-undiscovered
objects. When the discovery of Uranus made Herschel a famous man, it was
met with acclaim rather than skepticism.


As astronomers began to have a sense of the relative distances between the

planets, they noticed a surprisingly large gap between Mars and Jupiter. By
the mid-eighteenth century, it was recognized that the distances to the plan-
ets followed a pattern:
a = 0.4 + 0.3 2m
where a is the semi-major axis, or average distance from the Sun, and m
is the set of nonnegative integers, plus negative infinity. Given this odd-
looking set of rules, the following distances result:
1 0.4 Mercury 0.39
0 0.7 Venus 0.72
1 1.0 Earth 1.00
2 1.6 Mars 1.52
3 2.8 ??
4 5.2 Jupiter 5.20
5 10.0 Saturn 9.54

This pattern, called Bodes law or the Titius-Bode relation after the
mathematicians who popularized it, matches planetary distances well, but
some astronomers wondered about the missing planet that should be at
2.8 astronomical unit (AU), where 1 AU is defined as the mean distance
between the Earth and Sun. The discovery of Uranus furthered the suspi-
cion that additional planets may be lurking unseen in the solar system. In
addition, Uranuss distance from the Sun is pretty well predicted by Bodes
law with m = 6. All of this evidence led a Hungarian astronomer, Baron
Franz Xaver Von Zach, to propose a systematic search for the missing
planet in 1800. He compiled a list of European astronomers who might join
the effort, intending to assign a search area to each one and nicknaming the
group the Celestial Police.
As the invitations were making their way through war-torn Central
Europe, however, a discovery was already being made. Giuseppe Piazzi, a

Sicilian astronomer, was mapping the sky to help comet hunters make more
precise measurements. While taking positions of stars in the constellation
Taurus on January 2, 1801, he noticed that one of the star positions had
changed from the previous night. Further observations on the third and
fourth nights confirmed his suspicions. The new object didnt show any evi-
dence of a coma or a tail, and its motion was more typical of a planets
motion than a comets. Nevertheless, Piazzi conservatively announced that
he had found a comet and hoped further observations would prove his
greater hopes correct. Unfortunately, an untimely illness confined Piazzi to
bed, and by the time he recovered this object had moved too close to the
Sun to be observable.
Piazzis observations were not sufficient to allow an orbit to be calculated,
given the types of orbital calculations available at the time. Worse still, that
meant that there was little chance of finding it when it again became observ-
able, except via luck. Because of the possibility that Piazzis object was the
missing planet, there was great interest in using the observations available to
calculate a position, or at least to narrow down the search area to as small as
possible. Carl Friedrich Gauss, a young mathematics prodigy from Hanover,
turned his attention to the problem and developed new methods for calcu-
lating orbits, many of which are still used today. He calculated a predicted
position for later in the year, and on the last day of 1801 Piazzis object was
recovered by Von Zach. The new positions allowed further improvements
and greater certainty in the objects orbit, showing that it could indeed be
considered a planet and that its orbit was where Bodes law predicted it
should be. A delighted Piazzi named his object Ceres Ferdinandaea after the
Roman-era patron deity of Sicily (Ceres) and the then-current king of Sicily,
Ferdinand. In time, the object became known simply as Ceres. As 1802
opened, therefore, astronomers were content that they had found the
missing fifth planet. They were in for a surprise, however.
In March 1802, Dr. Heinrich Wilhelm Olbers of Bremen intended to make
observations of Ceres in hopes of improving the orbit of the newly discovered
planet. Instead, only a few degrees away from where Ceres was supposed to
be, he found another object of roughly the same brightness. This new object
proved to have a very similar orbit to Ceres, except for an orbital plane very
different from the known planets. Olbers named this new planet Pallas, and
astronomers struggled to make sense of the situation. Within five years, two
additional objects were discovered. In recognition of Gausss contribution to
astronomy, he was allowed to name one of the planets, choosing the name
Vesta. By this time, however, it was clear that the four new planets were quite
different from the others. William Herschels observations of Ceres and Pallas
showed that they were many times smaller than Mercury or the Moon. Even
at the highest magnifications with his largest telescope, they still had starlike
appearances in contrast to the other planets, all of which showed measurable
disks. For this reason, Herschel suggested that Ceres and Pallas (and later
Juno and Vesta) really were a different class of objects, which he named aster-
oids for their starlike appearance.
Historical Background  19


It was Olbers who first suggested an origin for the asteroids, which had a
profound effect on popular culture, if not an enduring acceptance in sci-
ence. Noting the small sizes of the asteroids, he wondered if their origin
could be explained as fragments of a normal-sized object that was disrupted
or exploded. This led astronomers to wonder whether there were further
asteroids to discover. Although there is ample evidence against this origin
for the asteroids, its legacy remains in the creation story of Superman and
the debris left after the Death Star destroyed Alderaan.
Ironically, the best support available in the early nineteenth century for
the exploded-planet hypothesis was published in the decade before Ceres
would be discovered, and further studies would show that asteroids could
not have originated on a single disrupted body.
Stones and pieces of metal falling from the sky have been reported for
centuries. As of the late 1700s, these were thought to be rocks flung from
distant volcanoes (even as distant as the Moon!) or perhaps created in the
Earths atmosphere. More prevalent still was the conclusion that the reports
themselves were mere folktales not to be taken seriously.
In 1794, however, Ernst Cladni published a book that reassessed these
reports and proposed that they could best be explained as due to objects
entering our atmosphere from outer space, often creating fireballs during
entry. Cladnis case was bolstered by the newly emerging disciplines of geol-
ogy and chemistry, and the lucky occurrence of several falls over the next dec-
ade. Chemical analyses of the fallen irons and stones showed that they both
contained metallic nickel, which is not seen in terrestrial rocks. The discovery
of Ceres, Pallas, Juno, and Vesta provided further evidence that small objects
could be present in-between the planets, though it was over a century before
asteroids and meteorites were conclusively associated with each other.


Just as Pallas was discovered by someone looking for Ceres, the fifth known
asteroid, Astraea, was discovered in 1845 by someone looking for Vesta, the
fourth. This discovery, and the discovery of three more asteroids in 1847, led
astronomers to reconsider the status of objects in the asteroid belt. The 39-
year gap between the discoveries of Vesta and Astraea would be the longest
drought for asteroid hunters, and Figure 2.2 shows how the number of known
asteroids grew quickly through the second half of the nineteenth century.
This increase in discoveries was accompanied by an appreciation of the
potential usefulness of asteroid studies by some mathematicians. By 1870,
Kirkwood had shown that the orbits of the 100 known asteroids spanned a
relatively wide range of distances from the Sun. However, the range of solar
distances covered by the asteroids was not uniform and avoided regions
where Jupiters pull was disproportionately strong because of mean motion
resonances, where the orbital period of an asteroid was exactly one-half,

Figure 2.2 Over the course of the middle and late nineteenth century, the number of
known asteroids mushroomed from four to more than 450. The pace of discovery was
relatively slow at first, but the introduction of photography led to an explosion of find-
ings, both via dedicated searches and accidental observations by astronomers studying
stars and galaxies. Illustration by Jeff Dixon.

one-third, or two-fifths the period of Jupiter (among other similar frac-

tions). Kirkwood argued that this was inconsistent with their origin in a sin-
gle destroyed planet, but was what one might expect if the planets formed
in a disk of gas and dust. This origin scenario is similar to the currently
accepted theories for the formation of the solar system (see Chapter 5).
By 1870, astronomers were also starting to take advantage of the new
technology of photography. Henry Draper and others first took pictures of
the Moon and Sun, following with pictures of stars. In the 1880s, Draper
took the first successful picture of a comet, and Max Wolf was the first to
use photography to discover an asteroid. Wolf went on to discover over 200
asteroids beginning in the 1890s. The use of photography enabled a perma-
nent record to be made of observations, with a much higher accuracy than
was available from sketches made by observers looking through telescopes
with their eyes. Photography was also combined with the new technique of
spectroscopy, where the targets light is split into its constituent colors,
which vary depending on composition, as described in greater detail in
Chapter 7. Spectroscopy of comets showed that they have organic material
(containing both carbon and hydrogen) on their surfaces.


In the closing decades of the nineteenth century, astronomers used their

newly developed tools like spectroscopy and photography and their new large
Historical Background  21

telescopes to study the universe outside of the solar system, which was avail-
able to them for the first time. As a result, interest in small bodies research
lagged. Indeed, as the new field of astrophysics was maturing, asteroids in
particular were often considered only as unwanted objects that littered their
photos of nebulae and galaxies and were sometimes derided as vermin of
the skies. This attitude was compounded by the small sizes of the newly
found asteroids. Furthermore, while the earliest estimates of the sizes of aster-
oids were typically overestimates, the new technique of photometry led to
large underestimates of their diameters. Photometric observations measure
the brightness of objects, which is a function of their distance as well as the
fraction of light they reflect, or albedo. The albedos assumed for the large
asteroids were poor estimates, and the sizes were wrong by a factor of two. By
this time, the term minor planet began to be used for the asteroids and com-
ets. Ceres, Pallas, and Vesta, with similar orbits to the hordes of new discov-
eries and deceptively small size estimates, were seen as much more similar to
their smaller cousins than Mars or Mercury. By the second half of the 1800s,
the largest asteroids had lost their status as planets.
However, some astronomers and mathematicians still found asteroids
and comets interesting because of their orbits. In 1898, 433 Eros was discov-
ered and was of immediate interest because of its proximity to Earth and
status as the first near-Earth object (NEO). It was realized that observa-
tions of Eros could take advantage of the same principle of parallax used by
Tycho in the sixteenth century. The technology available in the early 1900s
allowed Eross parallax to be measured, providing a value for the distance
between Earth and the asteroid, which was in turn used to calculate the dis-
tance between the Earth and the Sun. This was the most accurate measure
of that distance, known as the AU, available for many years.
The discovery of Eros also inspired two advances that have come to be
routinely applied to small bodies to this day. The first was the finding that
Eross brightness varied with time in a repeating pattern. It was quickly real-
ized that this brightness was most likely due to Eross rotation and that by
observing the change in brightness, or lightcurve, Eross shape and the
direction of its rotation axis could be calculated. Lightcurve studies have
been mainstays of asteroid and comet observations ever since.
The second advance was spurred by the desire to refine Eross orbit. Vast
collections of photographs of the sky were kept at observatories throughout
the world. After Eross orbit was first calculated, dynamicists realized that it
was probably present on some of these old photographs, and located its
position. These precoveries have been used to improve NEO orbits, par-
ticularly in cases where objects are potentially on a collision course with the
Within a few years of the discovery of the first NEO, a discovery was
made on the other side of the asteroid belt. Among the hundreds of aster-
oids found photographically by Max Wolf was an object that shares an
orbit with Jupiter. This object, 588 Achilles, orbits near one of the Lagran-
gian points of Jupiter. Objects in those orbits are protected from

perturbations by Jupiters gravity. Achilles was only the first of thousands of

objects, collectively called Trojan asteroids, now known in one of these
orbits. Mars and Neptune are also known to have objects at their Lagrange
The increased pace of discoveries to close the 1800s meant that studies
could be done on the entire population of asteroids, looking for trends and
patterns. This type of study was first done by Kirkwood, mentioned previ-
ously. In 1918, the Japanese astronomer Kiyutsugu Hirayama identified
close orbital similarities between some groups of asteroids. He correctly
deduced that these groups, which he called families, could have formed
from the breakup of objects with the fragments remaining in similar orbits.
These groups are now called dynamical families, or Hirayama families
after their discoverer. Dynamical families are still under study today, as they
provide unique opportunities to observe the effects of disruption and how
the surfaces of asteroids change with time.


With the increase of asteroid discoveries came an increase in the number of

orbits to catalog. In order to determine which observations were discoveries
and which were of already known objects, orbital calculations were required
for each known asteroid. The Astronomical Calculation Institute in Berlin
began this work in the 1870s, cataloging the known observations and add-
ing new observations to try and recover objects that had already been lost
(an ongoing project that stretched into the year 2000). The Berlin group
was also the first to publish alerts soliciting observations so that newly
found objects would not also join the ranks of the lost.
This work, interrupted by World War I, was officially sanctioned by the
new International Astronomical Union (IAU), which was formed in 1919
by a coalition of different national groups of astronomers. After World War
II, the IAU reassigned the work that had been done in Germany, with
astronomers in the Soviet Union responsible for publication of up-to-date
asteroid orbits, and the Minor Planet Center in Cincinnati responsible for
the alerts. The Minor Planet Center moved to Cambridge, Massachusetts,
in the 1970s, but still publishes Minor Planet Circulars via e-mail with the
latest discoveries and calls for observations.


At roughly the same time as the discovery of Astraea, astronomers noted dis-
crepancies in the position of Uranus. That planet had completed most of an
orbit since its discovery. More detailed calculations of its orbit were made to
try and understand these discrepancies, taking into account the effects of
Historical Background  23

Jupiter and Saturn. The French astronomer and mathematician Urbain Le

Verrier concluded that an additional planet must exist and was perturbing
Uranuss orbit. Le Verrier predicted a position for the new planet, and it was
later found close to the predicted position. This planet was eventually named
Neptune. Le Verriers achievement was much celebrated; it was the first suc-
cessful prediction of a planet based on theoretical principles, as opposed to
the accidental discoveries of Uranus and the asteroids.
Some 60 years later, there appeared to still be small discrepancies in the
positions of Uranus. Percival Lowell, founder of Lowell Observatory, used
the same mathematical techniques to predict the position of the missing
planet he felt must be responsible. He called it Planet X and led unsuc-
cessful searches until shortly before his death in 1916.
Even after Lowells death, the observatory he founded continued his
search. In 1929, the task of observing was given over to Clyde Tombaugh, a
young researcher from Kansas who had just begun work at the observatory.
Tombaugh began a systematic search of the skies for Planet X, and within a
year had discovered a bright, distant object at about the position Lowell
had predicted for Planet X. After receiving suggestions for its name from
the public, it was decided to name the newly found object Pluto.
Tombaugh and other astronomers at Lowell Observatory continued their
search through 1943, in the end covered roughly two-thirds of the sky
everything visible from their site in Flagstaff, Arizona. While Tombaugh
discovered more than 700 asteroids in the course of the search, Pluto was
the only outer solar system object discovered by the Lowell survey, which
seemed to confirm its status as a true planet.
As scientists learned more about Pluto, however, it seemed less like the
large planet it was originally thought to be. As discussed in later chapters,
astronomers were able to rule out a large atmosphere for Pluto, which was
puzzling. Other calculations showed that it should be much brighter than it
is if it were Earth-sized (or even Mars-sized). As touched upon in other
chapters, it was recognized that Plutos gravity could not be responsible for
the discrepancies in Neptunes orbit, and further planet searches were
made. Eventually, with the Voyager flybys of Uranus and Neptune in the
1980s, more precise values for the masses of these planets became available
and knowledge of their orbits were improved. Ironically, using the updated
masses of Uranus and Neptune in orbital calculations removed any discrep-
ancies in the orbit of NeptuneLowell and other mathematicians were
looking for a planet that didnt exist, and any similarity between Lowells
predicted position for Planet X and the location where Tombaugh found
Pluto was coincidence and luck!
Meanwhile, other astronomers began to expect other objects should be
found in the outer solar system. Gerard Kuiper theorized that a second aster-
oid belt might exist beyond Neptune, a belt of objects often called the Kuiper
belt objects (KBOs) in his honor. After years of searching, the first Kuiper
belt object was found in 1992. Hundreds of transneptunian objects (TNOs)

have been found since 1992, some of which are in the Kuiper belt, others of
which have different orbits, some of which are exceedingly similar to Pluto.
This work led directly to the demotion of Pluto by the IAU in 2006.
Small Bodies in Popular Culture
The small bodies of the solar system have ingrained themselves in the popular imagination. Repre-
sentations of comets have been found in the art of preliterate societies, which, given their interpre-
tation as omens, is not surprising. Comets remained striking emblems through the Dark Ages and
Renaissance; Giotto used a depiction of Comet Halley as the Star of Bethlehem in a painting in the
early 1300s (see Figure 2.3), and
the European Space Agency named
their mission to Comet Halley after
him. In the following centuries,
comets have been used as the name
for countless sports teams and
household products, along with
Bill Haleys backing band on Rock
Around the Clock.
The asteroids had a much later
start on entering the public con-
sciousness than did comets. By the
turn of the twentieth century, how-
ever, visits to asteroids became
commonplace in the newly evolv-
ing literature of science fiction.
The asteroids served metaphori-
cally as the setting for The Little
Prince, though there was no
attempt to be scientifically accurate
by author Antoine de Saint-
Exupery. As authors recognized
the potential of asteroids for both
Figure 2.3 The appearance of Comet Halley in European skies colonization and mining, asteroid
in the early 1300s inspired the Italian painter Giotto to colonies became settings for stories
include it as a representation of the Star of Bethlehem. inspired by the Gold Rush and the
Nearly 700 years later, the European Space Agency recog- old American West. On the other
nized Giottos cometary art by naming their first mission to hand, the potential hazard they
Comet Halley after him. Scala/Art Resource, New York. presented as impactors inspired
books and movies about the after-
math of collisions and how those collisions could be avoided. Misconceptions about the density of
the asteroid belt also led to dramatization of the danger of a ship passing through, perhaps most
notably in the movie The Empire Strikes Back, though also appearing in more abstract form in the
hugely successful video game Asteroids.
The most popular of the small bodies, however, is most certainly Pluto. Its discovery inspired the
name of a popular cartoon dog, which has helped its popularity, but its status as the furthest planet
has certainly been a factor. Perhaps oddly, the threat to its planetary status also added to its popular-
ity, and in fact that popularity was used by some as an argument to maintain its status as a planet.
Historical Background  25


The last decades of the twentieth century found our society experiencing an
information revolution due to the ready availability of computers. This rev-
olution also had great influence in astronomy and small bodies studies in
particular. The introduction of charge-coupled devices (CCDs) was a major
step forward for observations. CCDs are an integral part of the digital cam-
eras available today, and allowed quantitative measurements to be made in
a way that was difficult or impossible with film photographs. Because CCDs
store digital data, precise brightnesses can be measured. In addition, digital
pictures never degrade with time or poor storage. Telescopes with CCD
cameras can observe much fainter objects than was possible with film,
extending the productivity of many observatories and small telescopes pre-
viously thought to be obsolete or nearing the end of their useful lives.
The availability of computers and digital data has also allowed automatic
discovery of asteroids and comets. Development of software for automatic
detection and discovery was spurred by and designed for the NEO surveys
that began work in the 1990s, and it was a critical factor in the skyrocketing
discovery rate that followed. However, such software has also been used by
projects whose primary goal is not NEO discovery. The creation of CCDs
with ever-larger areas has made all-sky surveys possible for very faint
objects. Several of these surveys have been undertaken at various wave-
lengths, including the Sloan Digital Sky Survey (SDSS) and the 2MASS sur-
vey from the ground, and the Infrared Astronomical Satellite (IRAS) survey
from space. These surveys were designed for studies of galaxies, quasars,
and other distant phenomena, but as astrophysicists of the nineteenth cen-
tury found, the survey fields were littered with asteroids. Again, because of
the availability of digital data and computers, the information about aster-
oids from these surveys was easily extracted and has been a boon to plane-
tary astronomers.
The computational power available to theorists and modelers also has
allowed great strides to be made in simulating the formation of the solar
system, and the dynamics and orbital evolution of objects. Furthermore,
many of the calculations necessary for studying the possible effects of an
impact into the Earth or another planet would have been impossible to do
without the computers of today. The same is also true of image processing,
with the techniques used in planetary geology often applied to software
developed to look at medical imaging or even to touch up photos taken for
advertising purposes.


The final recent development in small bodies studies we will discuss here
is the dispatch of spacecraft. The focus of the American and Soviet space

programs on the Moon and planets (in particular Mars and Venus) still
allowed some small bodies to be visited, usually as part of other missions.
In this way the Galileo spacecraft performed the first visit to an asteroid,
flying by 957 Gaspra and then 243 Ida en route to Jupiter. The first visits
to comets had been performed in the 1980s by a flotilla of spacecraft to
Comet Halley and Comet Giacobini-Zinner. While American and Soviet
missions were frequently flown, there have also been dedicated European
and Japanese missions to small bodies. Indeed, the Japanese Hayabusa
mission was the first intended sample return from an asteroid, and the
European Rosetta mission is en route to being the first lander on a comet.
Other nations are also beginning involvement in small bodies missions as
they canCanada is pursuing a mission to discover and track NEOs from
The data returned from space missions has been a boon to planetary sci-
entists. Close-up views of asteroid and comet surfaces have allowed us to
gain a better understanding of the processes that are occurring there, which
has translated into better knowledge of how they formed and how any haz-
ard they might pose could be addressed. The samples returned from the
Stardust spacecraft have already shown that our ideas about solar system
formation are in need of revision, even as only a tiny fraction of the samples
have been studied.


The study of asteroids, comets, and dwarf planets has been central to the
history of astronomy. From the pre-telescopic era when observations of
comets showed that the heavens underwent changes, to spurring the
improvement of orbital calculations in the 1700s, to the modern era where
scientists have recognized the critical role of asteroids and comets in the
extinction of species, small bodies research has contributed important ideas
and important results both for other astronomical fields and for culture as a


The information on this Web site recounts how comets were seen as ill portents:
This site has a more general history of cometary studies: http://www.vigyanprasar.
The history of asteroid studies is detailed at this Web site:
This Web site focuses on the demotion of Ceres, Pallas, Juno, and Vesta from planet
to asteroid:
The Orbits and Dynamics
of Small Bodies

This chapter will discuss the places in the solar system where populations of
small bodies can be found. Some of those populations are still speculative,
with varying amounts of evidence for their existence. Other populations are
unquestionably real. To understand the places that asteroids and comets are
found, we will need to understand how to characterize their paths around
the Sun, and the forces that affect them.


The paths followed by small bodies as they travel around the Sun are called
orbits. In the 1600s, Johannes Kepler showed that objects orbiting the Sun
travel in elliptical paths, and move most quickly when they are closest to
the Sun and more slowly the farther away from the Sun they are. In addi-
tion, he showed that the time it takes an object to travel around the Sun (or
its orbital period) was related in a predictable way to its average distance
from the Sun (or its semi-major axis length). The semi-major axis is one of
six orbital elements that are used to fully describe and uniquely distinguish
an orbit. Two other important orbital elements for our purposes are the
eccentricity, or how noncircular the orbit is, and inclination, or how
the plane of an objects orbit differs from the plane of the Earths orbit. The
plane of the Earths orbit is also called the ecliptic plane. The point in an
orbit closest to the Sun is called the perihelion (also called q), the farthest


point the aphelion (also called Q). The positions of these points relative to
the Sun can be calculated as follows:

where a is the semi-major axis of the orbit and e is the orbits eccentricity.
Semi-major axis is usually measured in astronomical units or AU, where 1
AU is equal to the mean distance from the Earth to the Sun. Inclination is
measured in degrees; eccentricity has no units.
Isaac Newton generalized the findings of Kepler and solved the equations
of motion for any two objects moving only under the influence of each
others gravity. More complicated arrangements, such as for three or more
bodies, do not have simple solutions and, therefore, require numerical
approximations. Newton found that all matter will exert some gravity, and
that every object has a pull on every other object in the universe. This pull
depends strongly on the mass of the object, and even more strongly on the
distance to the object. For most objects in the solar system, the pull of
the Sun dwarfs the pull from every other object, and everything other than
the Sun can be ignored. For planetary satellites, the parent planet has a
much stronger pull than the Sun, and the Sun is usually ignored.


After the Sun, the most massive object in the solar system is Jupiter.
Jupiters gravity has had a profound effect on the orbital distribution of the
asteroids and comets throughout solar system history. The other planets
can also affect small body orbits. There are two main ways in which the
planets change the orbits of comets and asteroids: close encounters and
Close encounters are relatively straightforward. Under normal circum-
stances, an object will orbit the Sun, with the gravitational force of every-
thing other than the Sun unimportant. If a body happens to pass very close
to a planet (say, Jupiter, for instance), that planets gravitational pull on the
body can become comparable in size to the pull of the Sun. The path of the
body would then be influenced by both the Sun and the planet. The closer
to the planet the body passes, the more its orbit can change. This type of
orbit change is often purposefully used by spacecraft designers, for instance
the MESSENGER mission used two encounters with Venus to direct that
spacecraft to Mercury, and the Voyager 2 mission took advantage of a rare
alignment of the planets to be deflected by Jupiter to Saturn, by Saturn to
Uranus, and then by Uranus to Neptune. While Jupiter is most effective at
changing small body orbits, any planet can do it with a close enough pass.
The asteroid 99942 Apophis will pass very close to the Earth in 2029, and
the Earths gravity will significantly change that asteroids orbit. If it were
possible to make perfectly accurate measurements, we would find that the
The Orbits and Dynamics of Small Bodies  29

Earths orbit also changed as a result of the encounter, ever so slightly. This
is because the energy taken to change the orbit of Apophis around the Sun
came from the orbit of the Earth around the Sun. However, because the
Earth is so much more massive than Apophis, the change in Earths orbit is
not measurable.
The gravity of a planet can have much longer-term effects than during
brief close encounters. If one graphed the orbital period of all of the known
asteroids, there would be a distinct lack of objects with periods of 5.93 years
and 3.95 years, though many other bodies have similar periods. The missing
periods are related to the period of Jupiter, 11.86 years. If an asteroid had a
period of 5.93 years, its closest approach to Jupiter would always occur at
the same point in its orbit. Therefore, an extra force due to the gravity of
Jupiter will pull the asteroid in a particular direction. This steadily changes
the asteroids orbit, increasing its eccentricity. An object with a period of
3.95 years would approach Jupiter at only two points in its orbit. The effect
is not as strong on such an asteroid as one with a period of 5.93 years, but
will still have a large effect given the time available since the solar system
formed. This effect, called a mean-motion resonance, is important for many
solar system bodies. These two resonances are called the 2:1 (Jupiters pe-
riod is twice that of the hypothetical asteroid) and 3:1 (Jupiters period is
three times the asteroids). In the cases described here, the orbit of an object
is changed under the influence of a planet. Some mean-motion resonances
can, conversely, force an object to stay in an unusual orbit or serve as protec-
tion of a sort. For example, Pluto is in a 2:3 resonance with Neptune (it com-
pletes two orbits in the time it takes Neptune to complete three). This means
that it always has its closest approach to Neptune in the same part of its
orbit. However, Plutos orbit is quite eccentric, and this close approach
happens in the part of the orbit where Pluto is rather far from the Sun and
Neptune. Plutos orbit is so eccentric that it is sometimes closer to the Sun than
Neptune, and therefore their orbits cross. However, at that time, Neptune is
always far from Pluto. As a result of this resonance, Pluto and Neptune are
never particularly close together, and Plutos orbit is in no danger.
Another, more complicated type of resonance is called a secular reso-
nance. In a secular resonance, the precession of two bodies occurs in tan-
dem. This type of resonance is quite important for asteroids, and like the
mean-motion resonances, secular resonances have led to the ejection of
many objects from the main asteroid belt. The n6 resonance with Saturn in
particular has a major effect on the asteroidal population.


The force of gravity alone is sufficient to explain and calculate the motions
of most of the bodies in the current solar system (and, indeed, the uni-
verse). For small solar system bodies, however, nongravitational forces can

become important and must be taken into account when predicting their
motions over long periods. Ultimately, these other forces all derive their
power from the Sun, though sometimes in indirect ways.
The tails and comae that distinguish comets from other objects carry gas
and dust from the comets nucleus (see Chapter 10). They also carry mo-
mentum from the nucleus, which leads to a change in the comets orbit,
much like a spacecraft orbit is changed when thrusters are fired. As the
comet approaches the Sun and heats up, the rate of ice sublimation
increases, as does the amount of nongravitational force on the comet. The
exact change is dependent upon factors such as nucleus temperature, the
composition of the ices in the nucleus, and the spin rate of the nucleus,
among others. This makes it hard to predict the amount of nongravitational
force that a particular comet will experience. However, measuring the dif-
ference between a comets real orbit and the orbit expected with no nongra-
vitational force can give a measure of the nongravitational force it has
already experienced.
The Sun has a more straightforward connection with thermal forces.
Matter takes time to heat up and cool down. This property is called thermal
inertia. High thermal inertia material, like rock, takes longer to heat up and
cool down than low thermal inertia material like sand. This property is part
of the reason that the hottest part of the day, on Earth, is during the after-
noon, rather than at noonthe ground takes time to heat up. The diurnal
Yarkovsky effect is also due to this property. Light absorbed from the Sun
carries a small amount of momentum with it. The force from this solar
heating is symmetrical around the place directly beneath the Sun, where it
is noon. Reradiated heat from the object is symmetrical around the hottest
place on an object, where it is afternoon. The difference in these two direc-
tions puts a tiny torque on the objects orbit, changing it. Whether the orbit
speed increases or decreases due to the Yarkovsky effect is dependent upon
the direction of the objects rotation. This effect is very, very small, but for
110 km objects that have been going around the Sun for billions of years,
it is measurable and important. Depending on the direction of rotation, this
effect can either enlarge or shrink the objects orbit.
For objects with eccentric orbits, the seasonal Yarkovsky effect also occurs
during the course of a year. In this case, the effect is because the warmest
part of the year occurs at a different time than the time of the longest day
(this is seen on the Earth, where the longest day in the northern hemisphere
is in June, but the warmest part of the year is often in July or August). Here
again, the directions of average incoming sunlight and outgoing thermal
reradiation are not aligned, which changes the orbit. In the seasonal effect,
the orbit always changes so that it gets smaller with time. As with the come-
tary nongravitational forces, the size of the effect is dependent upon specific
properties of the object such as spin pole direction and spin rate, albedo or
fraction of light reflected, thermal inertia, and size. The Yarkovsky effect
was originally discovered in the early twentieth century, but was forgotten
The Orbits and Dynamics of Small Bodies  31

for decades before its rediscovery in the 1990s. It has been measured by pre-
cise radar observations of the asteroid 6489 Golevka, which moved 15 km
due to this force over the course of 12 years.
Once objects are small enough, heat conducts efficiently and rapidly
through them, so they are the same temperature throughout (or isother-
mal). The Yarkovsky effect does not occur on isothermal bodies. However,
at this size range other forces can become important. Poynting-Robertson
drag is similar to the seasonal Yarkovsky effect, but occurs for dust-sized,
isothermal objects even if they are in circular orbits. In Poynting-Robertson
drag, dust absorbs solar radiation and reradiates it. Einsteins laws of relativ-
ity show that the reradiation from the point of view of the Sun is not equal
in all directions. This, again, changes the dust grains orbit, making it
smaller and having the effect of spiraling the dust grain into the Sun. At
even smaller sizes, radiation pressure becomes most important. At these
sizes, roughly 1 mm or smaller, the momentum carried by light itself pushes
the dust grains away from the Sun, ultimately pushing them out of the solar
system altogether.


Given the large number of small bodies, and the forces that can act to
change their orbits, it is no surprise they are found throughout the solar
system. The following reservoirs are where most small bodies are found, or
are hypothesized. We begin far from the Sun and travel inward.

Oort Cloud

At a distance of tens of thousands of AU, a good distance to the nearest star,

we expect to find a large spherical shell of icy objects. At these distances, the
Sun is only as bright as Mars or Jupiter appear at their brightest in the
Earths sky, and it can take upward of a million Earth years to complete an
orbit. The Suns gravity is quite feeble that far away, and it does not take
much of a perturbationa passing star, perhaps, or a close pass between
bodies, to rip one of these objects from the Suns grip altogether, or send it
hurtling inward, after which interactions with the planets may make it a
more regular visitor. This shell, commonly called the Oort cloud after as-
tronomer Jan Oort, has never been directly seen, but astronomers are confi-
dent it exists. Why is this?
Oort and other astronomers noticed a pattern when looking at the orbits
of comets: comets with orbital periods of roughly 200 years or less have
orbits close to the plane of the planets; and comets with longer periods have
orbits that can be very far from the ecliptic plane, and come from random
directions. In addition, calculations of the dynamical lifetime of long

period comets, that is, the typical length of time before a long period comet
gets removed from the population either by striking an object or being per-
turbed out of the solar system, show that long period comets only last for a
few tens of millions of years, compared to the 4.5 billion years that the solar
system has existed. This implies a source where objects are still becoming
long period comets. The fact that they have random inclinations suggests
they are in a sphere centered on the Sun, and the size of the semi-major axes
of their orbits gives an idea of how far that sphere is.
It is believed that there as many as a trillion objects in the Oort cloud,
beyond 20,000 AU, with diameters of a few km. Objects still in the Oort
cloud are too faint to be seen, so this number is very, very uncertain. We
also do not know where the inner edge of the Oort cloud is, since objects
closer in than ~20,000 AU are more difficult to perturb by passing stars,
and as a result they end up in the inner solar system less frequently. Without
their orbits to trace back, the existence of the Inner Oort cloud is still a
matter of debate. Ironically, we know more about the structure of the region
farthest from the Sun. Simulations suggest, however, that the Oort cloud
does not extend closer than a few thousand AU. The known contents of the
vast space between a few thousand AU and a few hundred AU are only a few
objects, though it could be much more. The most notable object is 90377
Sedna, which does not come closer to the Sun than 76 AU, but retreats to
nearly 1000 AU. There is disagreement over whether Sedna is an Oort cloud
object that has been perturbed inward or an object from the Kuiper belt
(see following section) that has been perturbed outward.

Transneptunian Region

The next grouping is the transneptunian objects, or TNOs. The region they
inhabit comprises the space from the distance of Neptune (roughly 30 AU)
to a few hundred AU. The TNOs are icy bodies beyond Neptune, and are
thought to be similar to comets. Indeed, the short period comets (see fol-
lowing information) are thought to originate as objects from this region.
The TNOs are divided into several categories, dependent upon their
orbits. A large number of them are found beyond Neptune in orbits near
the ecliptic plane. This collection of objects is called the Kuiper belt (some-
times the Edgeworth-Kuiper belt), named after the Dutch-American as-
tronomer who proposed its existence. The first Kuiper belt objects were
found in 1992; hundreds are now known. Calculations suggest that roughly
70,000 objects larger than 100 km in diameter are present in the Kuiper belt.
Interestingly, there appears to be a drop-off in the number of objects with
semi-major axes more than 4550 AU. This Kuiper cliff is still being
investigated, but all indications are that there is a real lack of distant KBOs
and not an observational effect due to the increasing faintness of those
The Orbits and Dynamics of Small Bodies  33

Another group of TNOs are found in resonant orbits with Neptune. Of

these, most are in the 2:3 resonance, orbiting two times for every three Nep-
tune orbits. As mentioned before, the dwarf planet Pluto is among these
objects. For this reason, some call the smaller objects in the 2:3 resonance
plutinos, though this name is not an official one. Other objects are found
in the 1:2 resonance with Neptune, and there are a few objects known to be
in a 1:1 resonance (also called a Trojan resonance, see following section)
with Neptune. All of these objects are currently thought to have been cap-
tured into resonant orbits early in solar system history.
Finally, some TNOs are in orbits that suggest they have had close
encounters with planets, which have affected their paths. These can have
high inclinations and large eccentricities, but are not in resonances. These
bodies are called scattered-disk objects, and include the dwarf planet Eris.


Another group thought to be related to TNOs but orbiting closer to the Sun
are the centaurs. These objects have orbits that cross the orbits of the giant
planets, but are not in resonance with them. It has been suggested that the
centaurs are similar to the scattered-disk objects, but they have simply been
perturbed inward instead of outward. Some, including 2060 Chiron, the first
one discovered, have shown evidence of a coma. This has resulted in some
centaurs being listed as both comets and asteroids. Because their orbits cross
those of planets, their dynamical lifetimes are short. While most of the
objects discussed so far have likely been in their orbits for billions of years,
the current centaur population is transient, and they have likely only been
in their orbits for a few million yearsonly a few tenths of a percent as long
as the other populations. Numerical calculations also suggest they will only
last about that much longer before hitting a planet or being flung from the
solar system altogether, while new objects replenish the population.
This part of the solar system is where Comet Halley spends most of its
time. Its orbit suggests that it was originally from the Oort cloud rather
than the Kuiper belt, but Comet Halleys orbit has been changed by close
encounters with the outer planets so that it no longer retreats as far from
the Sun at aphelion. Other comets with similar orbits are called Halley
family comets. A distinction is also made for comets with periods longer
than 200 years versus those with shorter periods. Not surprisingly, those
objects that take longer than 200 years to orbit the Sun are called long
period comets, while the others are called short period comets. Long
period comets have semi-major axes beyond Neptunes orbit, while the
short period comets have semi-major axes near or within Neptunes orbit.
While the Halley family comets have semi-major axes that place them in the
outer solar system most of the time, their perihelia are in the inner solar
system, often interior to the Earths orbit.

It is useful, in classifying comets, to use the Tisserand parameter, a num-

ber that describes an orbit. This number is calculated differently with
respect to each planet. For Jupiter, it can be calculated as:

Where a, e and i are the semi-major axis, eccentricity and inclination

of the small bodys orbit, respectively, and aj is Jupiters semi-major axis
(5.2 AU).
The Tisserand parameter doesnt change even after a close passage by
Jupiter. If an encounter with Jupiter changes an objects semi-major axis, its
eccentricity will also change so that Tj stays the same. Halley family comets
typically have Tisserand parameters of less than 2 and periods of less than
200 years.


The orbit of Jupiter begins the realm of the asteroids. The first group of
asteroids we encounter as we move toward the Sun orbit at the same
distance as Jupiter (5.2 AU). Thus, these objects are in a 1:1 resonance with
Jupiter. Objects either leading or trailing Jupiter (or any planet) by 60 on
average can remain there without being perturbed. This was first determined
to be true by Joseph Lagrange, a French physicist, in 1772. In his honor,
these points (as well as three other equilibrium points) are called Lagrangian
points. The asteroids at the leading and trailing Lagrangian points of Jupiter
(also called L4 and L5) are called the Trojan asteroids. Asteroids discovered
and determined to be members of this group are typically named after
participants in the Trojan Wars. Trojan asteroids do not need to be exactly
at the 60 orbit difference from the planet, but from the point of view of
Jupiter they move in paths (or librate) around the Lagrange points.
The population of objects in the Trojan region is thought to be quite
large. The number of objects larger than 1 km is estimated to be several
hundred thousand, with a total mass about 10,000 times less than the Earth.
There is not much known about the Trojan asteroids besides their orbits. As
will be discussed elsewhere, details of their origin and compositions are still
under debate. However, it is thought that the Trojans are transitional
objects, somewhere between asteroid and comet. Trojan asteroids can have
very high inclinations compared to a typical main-belt asteroid. We do not
know of any Trojan-type asteroids around planets other than Jupiter, Mars,
or Neptune. Dynamical modeling suggests that Trojan asteroids of Saturn
and Uranus would not be stable for long periods of time. Any possible
Earth Trojans would be very difficult to detect from the ground, and
the searches that have been done to date have been negative. Venus and
Mercury Trojans are also thought to be dynamically unlikely.
Between the Trojan asteroids and the main asteroid belt are the Hilda
Group, at the 2:3 resonance with Jupiter. The Hildas can be seen as having
The Orbits and Dynamics of Small Bodies  35

Table 3.1. Major Asteroid/Comet Groups

Oort cloud >10000 Yes 0 >1 trillion?
Kuiper belt 3050 Yes, in general >800 100 million10
HFC 1535 No 50 ?
Centaurs 1030 No 100 10 million
Trojans 5.2 Yes >2000 300,000
Hilda 3.9 Yes 1500 25,000
JFC 27.5 No >300 100010,000
Main belt 1.93.4 Yes, in general >300,000 2 million
NEOs 0.92 No 4500 1100
Vulcanoids 0.090.2 ? 0 Hundreds?

the same relationship with Jupiter as the Plutinos do to Neptune. It is not

clear, however, whether the Hildas reached their current position from the
Trojan region or from the main belt, or indeed, if they formed near their
current orbits and are not derived from either group.
Jupiter family comets are those whose orbits are affected by Jupiter.
They are typically defined as having Tj between 2 and 3. It is thought that
Jupiter family comets originated in the Kuiper belt, as opposed to the Oort
cloud origins of the Halley family comets. Again, while their semi-major
axes place them near Jupiter most of the time, their perihelia can be near
the Earths orbit or even closer to the Sun.

Main Asteroid Belt

The main asteroid belt (often just called the main belt or asteroid belt)
spans the region between roughly 2 AU and 3.5 AU (sometimes the Hilda
region is included in the main belt, sometimes not). Figure 3.1 shows the
position of over 350,000 asteroids in terms of their semi-major axis and in-
clination. The region has noticeable concentrations of objects and areas
with few or no objects at all. These latter regions, called the Kirkwood gaps,
show the positions of some of the resonances mentioned before. The no-
ticeable gap near 2.5 AU, vertical on Figure 1, is where objects would be in a
3:1 mean motion resonance with Jupiter. The n6 secular resonance with
Saturn defines the inner boundary of the main belt at inclinations below
15 degrees, aided by the 4:1 resonance with Jupiter near 2.06 AU. The 5:2
resonance with Jupiter is also visible near 2.8 AU. There is an obvious cutoff
in the number of asteroids beyond 3.25 AU, near the position of the 2:1 res-
onance with Jupiter. Beyond that distance, the Cybele group orbit the Sun.

Figure 3.1The structure of the asteroid belt is shown by plotting the semi-major axis
and inclination of well over 300,000 asteroids. The vertical areas with few or no objects
are resonances with the outer planets. The Hilda group is near 4 AU and separated
from the main belt, as is the Hungaria group near 2 AU.

There are hundreds of thousands of known objects in the main belt.

Roughly two million objects with diameters 1 km and larger are believed to
be present in the asteroid belt. The size distribution is heavily weighted to-
ward the small end, however; less than 5,000 of those objects are thought to
be larger than 10 km.

Dynamical Families

Figure 3.2 shows a blowup of the middle of the asteroid belt. The Kirkwood
gaps are still obvious and visible, but in addition, clusters of objects are
obvious. These are dynamical families, first noticed by Japanese astronomer
Kiyotsugu Hirayama in 1918. The Koronis family (from about 2.82
2.95 AU and with i of about 2) and Themis family (i of 13 and
semi-major axis from 3.13.25 AU) are two of the families identified by
Hirayama shown in Figure 3.2. Families are named after their largest or
earliest-discovered member.
Dynamical families originate via collisions in the main belt. Collisions
can range from relatively minor events that create small amounts of ejecta
(or fragments created in the impact), to large cratering events that can expel
pieces of a meter in size or larger, to catastrophic impacts where the largest
The Orbits and Dynamics of Small Bodies  37

Figure 3.2When investigated in detail, the asteroid belt shows high concentrations of
objects with certain kinds of orbits. These clumps are dynamical families, objects on
similar orbits that originated through impacts onto, and often the disruption of, larger

remaining intact piece is less than half the mass of the original target (the
smaller object is always considered to be the impactor, the larger one the
target). The exact result of an impact is critically dependent upon the speed
and angle of impact, the relative sizes of the impactor and the target, and
the composition and strength of the bodies involved. Because asteroids and
comets are small, they do not retain most of the ejecta formed in an impact.
This is in contrast to the Moon, Mars, Earth, or other larger objects.
The extra speed that ejecta have, compared to the target, leads to slightly
different orbits. The differences between the speeds of the ejecta and parent
bodies are small compared to their orbital speeds around the Sun, however,
so the new orbits of the pieces are close to the original orbits. In some cases,
like the Karin cluster (a subfamily within the Koronis family, where a frag-
ment from that collision later suffered another collision), modeling has
allowed the orbits to be calculated backward in time, resulting in an esti-
mate of the time since the family forming event. For the Karin cluster, this
is a short timeonly between 5 and 6 million years. Most of the largest
familiesthe Eos, Koronis, Themis, and Flora families, for instanceare
thought to have originated billions of years ago, near the start of the solar
After a collision in the main belt, some of the new orbits for family mem-
bers or ejecta may place them in a resonance. In addition, some of the newly
created objects may be just the right size for the Yarkovsky effect to be a

factor in its motion. While initially the orbits of family members are very
similar, as time goes on the similarity can decrease and some objects can
have orbits that are quite different from their initial orbit and the orbits of
other family members. Another complicating factor is the existence of
objects that coincidentally have orbits similar to members of a family. These
objects, called interlopers, must be accounted for in detailed studies of
Dynamical families are also possible in other populations. The number
of objects expected in the Kuiper belt is certainly sufficient to allow for colli-
sions, although the collisional speed is much slower at those distances
(impacts occur at roughly 1 km/s in the Kuiper belt compared to 5 km/s in
the asteroid belt), and the number of known KBOs is much smaller than
the number of known asteroids. Nevertheless, it is likely that families in the
Kuiper belt will be identified in coming decades. There have been tentative
identifications of Trojan asteroid dynamical families, as well.
Just interior to the main belt, the Hungarias can be seen at inclinations
of 1825 degrees and semi-major axes of 1.82.0 AU. These objects are sepa-
rated from the main belt by a number of resonances, but are considered
part of the main belt by some.

Planet Crossers

Mars has its aphelion at 1.666 AU from the Sun. Objects that have perihelia
between 1.666 and 1.3 AU are called Mars crossers. These objects usually
have aphelia in the main asteroid belt, and are thought to have reached their
current orbits through a combination of weak resonances through the inner
belt and the Yarkovsky effect slowly affecting them over a long period of
time. The Mars crossers, like the centaurs, are not in orbits that are stable
for long periods, and their orbits will continue to change quickly until they
are removed from near-Mars space. In addition to these short-lived objects,
Mars is the only terrestrial planet known to have Trojan-type objects. Less
than 10 are currently known, 5261 Eureka being the largest with a diameter
of a few km. Mars Trojans appear to have orbits that are potentially stable
over the age of the solar system. Compositional studies, such as are detailed
in other chapters, suggest the Mars Trojans were captured into their current
orbits early in solar system history.
Small bodies with perihelia less than 1.3 AU from the Sun are classified
as near-Earth objects, or NEOs. These include both asteroids and comets.
These objects are of particular interest because some of them could poten-
tially strike the Earth at some point in the future. Indeed, the Earth has been
struck many times in the past, as described in more detail in the next chap-
ter. On the other hand, their proximity to Earth also makes them relatively
easy to visit with spacecraft compared to most other solar system objects. In
fact, some NEOs are easier to visit by spacecraft than the Moon is.
The Orbits and Dynamics of Small Bodies  39

Based on their orbits, NEOs are broken into further subclasses, shown in
Figure 3.3:
1. Amors, whose semi-major axis is between that of the Earth and Mars, and
whose closest approach to the Sun is outside the Earths orbit.
2. Apollos, whose semi-major axis is larger than the Earths, but whose closest
approach to the Sun is inside the Earths orbit.
3. Atens, whose semi-major axis is smaller than the Earths, but whose farthest
distance to the Sun is outside the Earths orbit.
4. Apoheles, whose semi-major axis is smaller than the Earths, and whose
farthest distance to the Sun is inside the Earths orbit.
Finally, there is another classification, that of potentially hazardous aster-
oids (PHA), which include all objects that have orbits that pass within
0.05 AU of the Earths orbit.
The Aten and Apollo objects in particular are of interest because their
orbits cross the Earths, making these the most immediately dangerous
bodies. However, the effects of close planetary encounters can change
NEO orbits, sometimes drastically, and objects can evolve from one sub-
category to another. Indeed, as discussed, all NEOs were originally from

Figure 3.3Near-Earth Objects are classified into different groups depending upon the
specifics of their orbits. Amors and Apoheles do not cross the Earths orbit but come
close to it from the outside and inside, respectively. Apollos and Atens both cross the
Earths orbit but have semi-major axes larger or smaller than the Earth, respectively.
The orbit of Mars is also shown, for comparison, along with typical orbits for these dif-
ferent NEO groups. Illustration by Jeff Dixon.

a different part of the solar system. The average time between becoming
an NEO and either impacting a planet or the Sun is only a few million
years, much shorter than the time since the solar system began. There is
a constant delivery of objects from the main belt via resonances and the
Yarkovsky effect to replenish the NEOs that are removed from the
The majority of known NEOs are in the Amor class. Very few Apoheles
are known. It is currently thought that this is due to a bias in the way aster-
oids are observed and discovered. Because Amors are further from the Sun
than the Earth is, there are times they can be observed all night long. Apoh-
eles (and to a lesser extent Atens and Apollos) all spend some time, or even
all of their time, in the daytime sky. This makes them very difficult to
observe. As a result, careful modeling based on the best data available is
required to come up with an accurate measure of how many Apoheles,
Atens, and Apollos exist. Over 4,000 NEOs of all sizes are known, and it is
believed that roughly 1,100 of them are larger than 1 km in diameter. About
a quarter of all NEOs are also PHAs. The proportion of comets to asteroids
in the NEO population is currently a matter of debate. Before the impor-
tance and efficiency of resonances was recognized, it was thought that
nearly half of the NEO population had to be cometary to account for their
numbers. More recent dynamical modeling in conjunction with composi-
tional studies places the fraction of cometary NEOs closer to 15 percent at
the most. The Tisserand parameter is often used to identify possible comet
candidates in the NEO population.


Vulcanoids are a hypothetical population of small bodies with orbits near

or interior to Mercury. They are named after Vulcan, a planet thought to
exist in the 1800s and early 1900s. This planet was thought to be necessary
to explain perturbations in the orbit of Mercury, in much the same way that
Neptunes existence was predicted from perturbations of Uranus. Some
observers even claimed to have seen transits of Vulcan across the solar disk.
None of these observations were ever confirmed, however, and with
Einsteins formulation of relativity, the need for an extra planet to explain
Mercurys orbit disappeared.
Vulcanoids, if present, would represent leftover unaccreted bodies from
the earliest solar system times. Searches using high-altitude aircraft and so-
lar telescopes have found no objects interior to Mercury larger than 50 km,
and the Yarkovsky effect would be expected to remove small objects effi-
ciently. However, there is still a possibility that vulcanoids between these
sizes could be found, because stable orbits are known to exist interior to
Mercury down to 0.09 AU.
The Orbits and Dynamics of Small Bodies  41


There are four main areas where asteroids, comets, and dwarf planets are
found: the main asteroid belt between Mars and Jupiter, the Trojan clouds
leading and trailing Jupiter, the Kuiper belt beyond Neptune, and the Oort
cloud stretching many thousands of times further from the Sun than the
Earths orbit. However, in addition to these stable orbits, small bodies can
be found in orbits that are unstable over the span of tens of millions of
yearsthese are the centaurs and near-Earth objects, among other smaller
populations. The existence of objects in these unstable orbits means that
bodies must move from stable regions to refill the unstable regions. It is
thought that this occurs via impacts and the Yarkovsky effect for asteroids,
while collisions and infrequent passages close to nearby stars change the
orbits of comets.


A flash-based online orbit simulator can be found at this Web site, including sev-
eral preset situations (such as a Trojan asteroid-like case) and the ability to
input positions and velocities to allow further exploration: http://phet.colorado.
This site has a more technical introduction to the mathematics behind orbital cal-
culations, including several further references:
The positions and orbits of comets, asteroids, and dwarf planets are provided by
the NASA Jet Propulsion Laboratory at this Web site. Views are available of the
orbits of individual objects, or positions of small bodies in large regions of the
solar system:
The discovery of Sedna, thought to be the only Oort cloud object currently known,
is detailed at this Web site:
For more information about further Hubble Space Telescope observations: http://
Meteors, Meteorites,
and Meteoroids

The Earth is under constant bombardment from a rain of extraterrestrial

material. In a typical year, the Earth is impacted by 54 tons of material.
Most of it is the size of dust grains, and does not penetrate Earths atmos-
phere. However, Earth is hit with objects the weight of a pencil or heavier
roughly 100 times a day. Objects the size of marbles burn up in the atmos-
phere and are responsible for meteors, or shooting stars. These can occur
either randomly or in periodic, predictable meteor showers.
Chair-sized and table-sized objects often strike the Earth after spectacular
fireballs, and fragments can survive to reach the ground as meteorites. Me-
teorite falls have occurred in Peekskill (New York), Monahans (Texas), and
Tagish Lake (Canada) in the last 15 years. Earth bears ample evidence of
even larger impacts. Every few decades, on average, Earth is impacted by an
object of 10 tons or sothe mass of the Hubble Space Telescope. A house-
sized piece of iron blasted a 1-kilometer-wide hole in the Arizona desert less
than 50,000 years ago. Such an event occurs, on average, every few tens of
thousands of years. An impact near the Washington Monument from an
object of 300 m in diameter would leave a crater that would stretch from
the Pentagon to the Capitol, with an ejecta blanket stretching halfway to
Baltimore and destruction resulting across the Eastern United States.
Craters dot the North American landscape from Texas to Virginia, from
Ontario to the Yucatan, as well as locations throughout the world. An
impact in Mexico is thought to have resulted in the extinction of a large


fraction of the Earths life 65 million years ago, including the dinosaurs. It
has been proposed that other so-called mass extinctions are also associated
with impacts. Luckily, such large impacts appear to be relatively rare, with
tens of millions of years between impacts of that size.
It is important to remember the difference between meteors and meteor-
ites. Meteors all burn up in the atmosphereone can never pick up a
meteor. If an object from space makes it to the ground, it is called a meteor-
ite. This is true regardless of the size of the pieces that land. These words
can be easily confused. Indeed, the most famous landmark associated with
an impact in the United States, Meteor Crater, is actually named incor-
rectly! A third concept with a similar name is that of meteoroids, small
objects in space. Normally they are considered to be smaller than asteroids,
but there is no official size range for asteroids or meteoroids. Before a mete-
orite strikes the ground, it could be called a meteoroid.


Meteors can be seen on most nights under sufficiently dark skies. Often
coming at a rate of a few per hour, and from random directions, these
meteor phenomena are due to the chance encounters of small objects with
Earth. These are called sporadic meteors. In contrast, several times a year
meteor showers occur. In a meteor shower, the rate of meteors goes up
from a few dozen to hundreds or even thousands per hour, and the paths
the meteors follow seem to come from a particular spot in the sky, called
the radiant. Meteor showers are named for the constellation in which their
radiant falls; for instance, the Geminids appear to come from Gemini, and
the Orionids from Orion. Meteor showers occur at the same time every
yearevery August 12th, the Perseids are at their peak, for example. How-
ever, the rate of meteors can vary substantially.
While the Perseids consistently provide about 100 meteors per hour near
their peak, the Leonids have varied from years when few meteors were seen
beyond the typical sporadic contribution, to years like 1833 (see Figure 4.1),
where over 100,000 meteors per hour were seen. That level of activity, often
called a meteor storm rather than a shower, translates to over 25 meteors
per second. More recent Leonid storms occurred in 1966 and around the
turn of the twenty-first century. These storms had a more modest rate of
only a few thousand meteors per hour, but they were unforgettable for
those who witnessed them.
The clouds of particles that cause meteor showers orbit the Sun, like every-
thing else in the solar system. We know the position at a specific time (for
instance, the Perseid meteors are at the Earths position relative to the Sun on
August 12th of every year), we know the direction of motion (out of the con-
stellation Perseus, in this case), and can measure their speed. When put to-
gether, the characteristics of the orbit can be calculated. In the case of the
Meteors, Meteorites, and Meteoroids  45

Figure 4.1This depiction of the 1833 Leonid shower, though produced over 50 years
later from secondhand accounts, gives a sense of the nature of meteor storms, with
dozens of meteors visible at all times. This meteor storm gave rise to the first scientific
explanation of meteors as extraterrestrial in origin rather than being simply weather-
related. NASA.

Perseids, the calculated orbit is very similar to that of a comet, Swift-Tuttle.

Indeed, in many cases meteor showers are associated with comets. It is for
this reason that, in general, meteors are thought to be derived from comets.

The Discovery of Meteor Showers
It was not until the Leonid storms of the 1830s that meteors first received real scientific attention.
Previously they had been thought to be atmospheric phenomena, and were named accordingly
(the word meteor has the same root word as meteorology, the study of weather). American astrono-
mer Denison Olmsted observed the 1833 storm from Connecticut and noted that all the meteors
appeared to emerge from a discernable region of the sky, the radiant. He collected reports of obser-
vations from throughout North America, and first proposed that meteors were coming from outer
space rather than from within Earths atmosphere. It was not long before the annual occurrence of
the Perseids was recognized, and other meteor showers also identified. (Interestingly, the Perseids
were identified by nonscientists centuries earlierthe shower was known to German peasants as
the tears of St. Lawrence, since it fell on the anniversary of his death.)


Meteorite discoveries are divided into two categories, falls and finds. Falls
are meteorites that are seen as fireballs and traced to their resting place. Very
rarely, this can be done without much effort, such as the handful of meteor-
ites that have struck houses (and even cars!) or fallen within the sight of peo-
ple. More often, after reports of a fireball, a search area can be defined and
fragments found. With video evidence, such as with the Tagish Lake fireball/
meteorite, searches can be done quickly and effectively. Other meteorites are
never found due to sketchy or contradictory eyewitness accounts of a fireballs
speed and direction. Falls are scientifically very interesting because they have
the best chance of retaining delicate minerals that are stable in the vacuum of
space but quickly degrade with exposure to Earths atmosphere.
Finds are meteorites that are found without evidence of a fireball. Unlike
falls, where the amount of time spent on Earth is known to be very short,
finds may have spent thousands of years or longer on Earth before being rec-
ognized as meteorites. Most meteorites in the worlds collections were found
in Antarctica, where the United States and Japan both send teams every year
for the purpose of collecting finds. There are several reasons Antarctica has
excellent meteorite hunting grounds. First, and perhaps, most obviously,
Antarctica has vast stretches of ice where almost any rock encountered has a

Figure 4.2 Meteorite searchers take advantage of a natural concentration mechanism to

increase their odds of finding meteorites in Antarctica. By studying the flow of glaciers,
certain areas have been identified where meteorites are carried long distances and then left
behind where the glaciers melt. Expeditions searching these areas have had excellent suc-
cess and typically bring back hundreds of meteorites per year. Illustration by Jeff Dixon.
Meteors, Meteorites, and Meteoroids  47

high likelihood of being a meteorite. Searches center on particular areas

where glaciers are known to be eroding at a rapid rate. Because glaciers move,
and because the ice erodes rapidly but rock doesnt, any meteorites that might
have fallen on the glacier are collected in one place.
Before the Antarctic expeditions began, most meteorites were found in
areas where the wind is removing soil, leaving any ancient meteorites
behind. The discoverers were often farmers, who were familiar with their
land and recognized any unusual rock types. To this day, nomads in the
Sahara desert find a significant number of meteorites for similar reasons.
Because these unusual rocks were not always recognized as scientifically
valuable, there are many stories of meteorites sitting on mantelpieces for
decades, or even being used as doorstops!
Meteorites are typically named for the town nearest their fall or find (in
populated areas, the nearest post office). For areas that have no nearby
towns, geographic features are used (for instance, Tagish Lake or even
Sahara). Antarctic meteorites come from a few specific areas and are num-
bered as well as named, with the number derived from the year of collection
and the order in which it was characterized. Hence, ALH 84001 was col-
lected in the Alan Hills of Antarctica in 1984 and was the first one to be

Recognizing Meteorites
Finding meteorites outside of a dedicated expedition is a very, very improbable event. However, it
does happen a handful of times each year. Meteorites have several characteristics that are unusual
compared to normal rocks: they have smooth sides and a coating called a fusion crust due to their
passage through the atmosphere; they usually contain at least some metal; and they do not contain
layers of any kind. There are resources on the Internet to help determine if any suspect rocks are
meteorites or meteor-wrongs. For those who prefer to leave the searching to others, some mete-
orites are common enough that they are sold at reasonable prices by reputable dealers.


Meteorites are classified into different groups defined by their composi-

tions. The first classification system was very general and included three
groups: stones, irons, and stony-irons. This level of classification requires
no special equipment, and requires no special training: stony meteorites
appear to be rocks, iron meteorites are hunks of metal, and stony-irons
contain both metal and rock. While most meteorites are thought to come
from asteroids, we know that some meteorites originate on the Moon and
Mars. Those meteorites will not be discussed in this book, but descrip-
tions of them and what we learn from them can be found in other

The original object that the meteorites came from is called the parent
body. Some parent bodies are thought to have been destroyed in a large
impact, with the meteorites found being all that remains of the original.
Other parent bodies are thought to still exist today (the Moon and Mars
and large asteroids like Vesta or Ceres are the most obvious examples), with
the meteorites being knocked off in relatively small collision events.
Through laboratory investigation of meteorites, scientists have found
variation within all three of these groups, and meteorite scientists (or mete-
oriticists) use a different scheme, separating meteorites into two major cate-
gories: chondrites (all of which are stony) and achondrites (including all
irons and stony-irons, as well as some stones).


Chondrites are so-called because they contain small, mostly spherical bits
of glass called chondrules. The abundance of chondrules can range from a
small amount all the way up to comprising most of the meteorite. The ori-
gin of chondrules is something of a mystery at present. They show evidence
of having reached high temperatures (over 1500C) in only a few minutes,
cooling in the course of a few hours. It is not known what caused the heat-
ing, with leading candidates including shocks, lightning in the solar nebula,
and outbursts from the early Sun, among other possibilities. Each candidate
has proponents, though none can yet explain all of the characteristics seen
in chondrules.
The part of a chondrite meteorite between the chondrules is called the
matrix and is composed of very fine-grained minerals. The chondrites are
thought to have survived largely unchanged since their formation roughly
4.5 billion years ago. When a mixture of minerals is heated to a given tem-
perature and pressure, a new set of minerals is formed in equilibrium with
one another, with compositions appropriate to the temperature and pres-
sure that was reached. The combination of minerals found in chondrites is
often unequilibrated, with no temperature and pressure conditions that
can explain all of the compositions of the minerals. This shows they have
not experienced high temperatures since the time these different minerals
were brought together. A period of high temperatures would have led the
minerals to exchange atoms and led the set of minerals to reach equilibrium
with one another.
Chondrites are divided into three major classes, the carbonaceous
chondrites, enstatite chondrites, and ordinary chondrites. (Ordinary
chondrites are named so because they are very common among meteor-
ites, not because they are uninteresting!) The ordinary chondrites are the
meteorites most commonly seen to fall, amounting to roughly 80 percent
of falls. The ordinary chondrites are further subdivided into the H, L,
and LL groups depending on the amount of iron they containH has
Meteors, Meteorites, and Meteoroids  49

the most, L less, and LL the least. The H chondrites are the most com-
mon of the ordinary chondrites, accounting for nearly 40 percent of all
falls. The L chondrites are somewhat less common, at roughly one-third
of all falls, and the LL chondrites account for slightly less than 10 percent
of all falls.
Carbonaceous chondrites were named because at the time they were
thought to contain more carbon than other meteorites, though more recent
studies have shown that, in fact, these meteorites do not always have more
carbon than other types of meteorites. Today, the presence of specific ele-
mental ratios is used to recognize members of this group. Like the ordinary
chondrites, the carbonaceous chondrite group is subdivided into several
subgroups: the CI, CM, CV, CO, and CR subgroups are the major divisions,
differing in metal content and the minerals that are present. Roughly 5 per-
cent of all falls are carbonaceous chondrites.
Two components of carbonaceous chondrites are worth particular notice
and are further evidence of their ancient nature. Whitish-colored specks
and grains are found in some carbonaceous chondrites. When analyzed,
these grains turn out to be rich in calcium and aluminum, giving rise to the
name calcium-aluminum-rich inclusions, or CAIs. The CAIs are the old-
est known objects in the solar system, with ages of 4.57 billion years since
formation, roughly 2 million years before the event that formed the
There are other small grains whose atomic composition is very different
from those seen in any other rocks, or from most of the grains in their host
meteorite. These are called presolar grains and are believed to have formed
in other stars and been ejected in supernova explosions or in outflows from
red giants. These grains, according to theory, randomly found themselves
near the Sun very early in solar system history. Most of them were heated
and destroyed, becoming homogenized with the rest of the material that
became the solar system, but a small number (luckily) escaped that fate and
were incorporated into meteorites. We do not know enough about the
grains or the candidate stars that may have been the original location of the
presolar grains in meteorites, but it seems possible that rocks in future star
systems could include grains that formed in our own Sun.
Unlike the carbonaceous chondrites, the name for the enstatite chon-
drites is an apt one, as these have a large fraction of the silicate mineral
enstatite. Enstatite is a type of pyroxene, one of the most common min-
eral groups on the Earth. Enstatite contains little or no iron and is very
rarely found on the Earths surface, but is predicted to be rather more
common in the Earths mantle. It has also been identified in disks of dust
orbiting other stars. Geochemists have determined that enstatite chon-
drites must have formed in water-poor and oxygen-poor parts of the solar
nebula, in contrast to the other chondrites, particularly the carbonaceous
chondrites. The enstatite chondrite group accounts for less than 2 percent
of all falls.


Meteorites without chondrules are called achondrites. These meteorites

have experienced a variety of heating and melting processes that have modi-
fied their compositions and original textures. The meteorites that have only
experienced a little melting are called primitive achondrites. More com-
monly found are pieces of objects that have been heated enough to separate
into a crust, mantle, and core, or differentiated.
Some iron meteorites are pieces of asteroidal cores, formed when their
parent bodies differentiated. These meteorites, called magmatic irons, give
insights into the conditions and composition at Earths core, for which we
have no samples and little hope of directly measuring. Brachinites are mete-
orites very rich in olivine, the most common mineral in the Earths mantle
(and called peridot when used as a precious stone). Olivine is detected
throughout the solar system and beyond and is composed of very common
elements: magnesium, iron, silicon, and oxygen. Because of the high con-
centration of olivine they have, brachinites are thought to be from aster-
oidal mantles. These are quite rare, perhaps because they are not strong
enough to survive collisional evolution in the asteroid belt and passage
through Earths atmosphere.
There are also several kinds of meteorites that represent crustal mate-
rial. The most common are the HED meteorites. This group (standing for
howardite, eucrite, diogenite) is thought to come from the asteroid
Vesta, providing one of the few well-accepted connections between a me-
teorite and parent body. The connection was formed by both geochemical
and astronomical argumentsthe reflectance spectrum of Vesta was
found to be a very close match to the spectra of HED meteorites obtained
in the laboratory (for further details of this technique, see Chapter 7). In-
dependently, geochemists studying the HED meteorites concluded they
had to originate from a large object that was likely still intact, and Vesta
was the best candidate. Further work studying main-belt and near-Earth
objects sharing Vestas spectral and orbital characteristics has strengthened
these conclusions.
Another group of melted meteorites are the angrites. Laboratory experi-
ments and theoretical calculations have been performed on the angrites and
HED meteorites (Vesta) to try and determine their compositions before
they were heated and melted. The results suggest that the parent bodies of
both of these meteorite groups had carbonaceous chondrite-like composi-
tions before they were melted.
Unlike the achondrites mentioned before, there are several achondrite
groups whose origins are not obvious and are still a matter of controversy.
The non-magmatic irons are thought to have formed through localized
impact melting of chondritic materials rather than via melting of the whole
object and differentiation. The stony-irons have two main groups, the mes-
osiderites and pallasites. Both are roughly equal mixtures of silicates and
Meteors, Meteorites, and Meteoroids  51

metal, with the pallasites often containing large crystals of olivine. Indeed,
the name of the mesosiderites reflects their general composition (meso
meaning intermediate or middle, while siderite is an older term for iron
meteorites). Pallasites, confusingly, have nothing to do with the asteroid
Pallas, but were named after the German naturalist Peter Pallas years before
the discovery of the asteroid that unrelatedly shares his name.
The formation of stony-iron meteorites is not well understood, but there
is general agreement that it was a complicated process involving melting
and differentiation followed by impacts with other objects while still


The smallest meteorites usually dont even make it to the ground. As men-
tioned previously, most of the mass that hits the Earth is in the form of
dust. Regardless of how dedicated or fastidious the search team, these inter-
planetary dust particles (or IDPs) will never be seen to fall or collected in
the usual way. Instead, collection is done by airplanes or balloons in the
stratosphere. Chemical analysis of IDPs is difficult, but they have been
roughly separated into at least two groups, hydrous and anhydrous,
depending on whether or not they contain water in their minerals. The
hydrous IDPs are similar to the CI carbonaceous chondrites, while the an-
hydrous IDPs are not similar to any meteorite type. This has led many sci-
entists to suspect that the anhydrous IDPs are cometary in origin, while the
hydrous ones are thought to be formed as the result of collisions between
asteroids. In addition, there are some IDPs that appear to be from other star
systems, similar to the presolar grains in carbonaceous chondrites.
The Ulysses spacecraft, a mission designed to study the Sun and solar
wind, found streams of dust moving through the solar system, and based
on the direction they were moving, scientists believe some of those streams
come from outside the solar system. The Stardust spacecraft, whose main
mission was to return samples of dust from a comet tail, flew through one
of those streams hoping to capture and return some samples to Earth, a
study whose results will become known in coming years. Further informa-
tion about Stardust and small bodies missions in general can be found in
Chapter 12.


If one could count every atom in the Sun, we would find that some ele-
ments are more numerous than others. Hydrogen is by far the most com-
mon element in the Sun, followed by helium. Every element down to
uranium would be found, however, to some degree. The relative amounts

of all of the atoms in the Sun are called solar abundances, some of which
are shown in Table 4.1.
When we look at the chondrites and measure the elements present, we
find that most of the elements are present in the same ratios as in the Sun
the relative amounts of magnesium and silicon, for instance, or of ytter-
bium and praseodymium. However, some elements are rarely found in the
chondrites compared to the amounts found in the Sun, like helium, simply
because those elements do not remain easily bound to rock. The distribu-
tions of elements in the chondrites are, not surprisingly, called chondritic
abundances, also shown in Table 4.1. The similarity between the chondritic
and solar abundances is consistent with the expectation that the bodies of
the solar system all started out made of the same components as the Sun,
modified for different temperatures and gravity in different parts of the solar
system. For instance, there is less nitrogen in the chondrites compared to the
Sun because nitrogen does not commonly occur in rock-forming minerals.
The major rock-forming minerals, like silicon, magnesium, iron, and alumi-
num, all occur in the same ratios in the chondrites and the Sun to within a
factor of two, which is considered an excellent match in this type of analysis.
Table 4.1. The Relative Amounts of the Elements in Chondrites


Hydrogen 28000 0.19
Carbon 10 0.32
Oxygen 24 4.6
Magnesium 1 0.93
Iron 0.9 1.8
Gold 0.0000003 0.0000013
Aluminum 0.082 0.082
Sodium 0.06 0.047
Nitrogen 3.1 0.03
Nickel 0.05 0.1

If we could count every atom that makes up Earth, we would expect to

find it to have chondritic abundances of the elements. However, the ele-
ments would not be evenly distributed through Earth. Most of Earths
nickel, for instance, is found in the core and very little is in the mantle or
crust. By contrast, most of Earths magnesium is in the mantle and crust,
with little in the core. Elements that tend to be found in the core are called
siderophile (metal loving), while those found in the crust and mantle are
called lithophile (rock loving).
Meteors, Meteorites, and Meteoroids  53


As you are probably aware, atoms consist of protons, neutrons, and elec-
trons. Elements are defined by the number of protons in their atomsif an
atom has one proton it is hydrogen, if it has two it is helium, if it has six it
is carbon, and so on. However, atoms from the same element can have
different numbers of neutrons, and thus different masses. These are called
isotopes. Carbon has three isotopes, with atomic weights of 12, 13, and 14,
symbolized like so: 12C, 13C, and 14C.
Isotopic studies of meteorites are used in several ways. Many isotopes
are radioactive, changing into stable isotopes of other elements after their
decay. For instance, 14C decays into 14N. Geochemists can study the iso-
topic mix of elements in minerals and calculate how long it has been since
they have formed. As an example, potassium (K) is an element commonly
found in rocks. One of the potassium isotopes, 40K, decays into argon
(40Ar) at a known rate. Argon does not form compounds with other ele-
ments, and is not found in minerals except as trapped gas molecules. So
by measuring the amount of 40Ar found in a rock, and measuring the
amount and kinds of K isotopes in the same rock, the age of the rock can
be calculated. It is on the basis of measurements like this using various ele-
ments that radiometric ages have been calculated for everything from
pottery shards and archeological sites to meteorites, lunar samples, and
the planets.
Stable isotopes have also been used to gain insights into meteorite
histories. The mix of oxygen isotopes16O, 17O, and 18Ois thought
to be fixed on an object since its formation. The relative amounts of
those three isotopes are often used as a fingerprint of a parent body.
Lunar meteorites were identified as such in part because they have the
same oxygen isotopic ratios as samples returned by the Apollo astro-
nauts. In turn, theories of the Moons origin must account for the fact
that the Moon and Earth share the same oxygen isotopic ratios and
must have had a common origin. Different meteorite types are grouped
together based on their isotopic ratios, and conversely those ratios can
be used to conclude that similar-seeming meteorite groups formed on
totally separate bodies.
Another important set of isotopes are those of hydrogen. Most hydrogen
has a proton but no neutrons. Deuterium is a form of hydrogen with both a
proton and a neutron. The hydrogen in the Earths water is roughly 0.015
percent deuterium. This fraction, sometimes called the D/H ratio, is also
used as a fingerprint. The D/H ratio in comets has been measured as signifi-
cantly higher than that on the Earth, suggesting that cometary impacts may
not have been an important source of the Earths water, or at the least that
water with a much lower D/H ratio must also have been brought from
another source or present on the early Earth in order to have the final ratio
we see today.


Meteorites provide a record of many different processes since the solar sys-
tem began. While dramatic changes are rare, especially in the last 4 billion
years, subtle change occurs to this day in the asteroid belt.
The most drastic changes seen in meteorites are the result of heating.
As the meteorite parent bodies were being formed, they accreted varying
amounts of radioactive elements. As those elements decayed, they gave
off heat. Larger objects retained more of this heat than smaller objects,
and those that formed more rapidly had a longer time to capture the
radiogenic heat than those that formed slowly. As a result, some large
objects got hot enough to melt. Starting with a chondritic composition,
small amounts of melting would create the primitive achondrites. As
temperatures continue to rise, the molten metal and silicates would begin
to separate because liquid silicates did not have the strength to support
the much denser metal. The separated metal would sink and pool with
nearby metal, accelerating the process. The sequence would end with the
metal near the center of the parent body as a core. Evidence from meteor-
ites suggests that this happened in dozens of parent bodies, perhaps as
many as 75100. This evidence comes not only from the iron meteorites
that obviously result, but also from the silicates that are left behind after
There is also ample evidence in the meteorites for thermal metamor-
phism. This is heating to temperatures short of the melting point for the
materials in the rock. As heating increases in chondrites, the minerals equi-
librate with one another, followed by changes in the grain sizes of minerals
in the matrix. In the final stages before melting, chondrules become indis-
tinct and begin to merge with the matrix.

Figure 4.3 This schematic shows how meteoritic materials change with heat. In this case, a
carbonaceous chondrite is formed from a mixture of rock and ice. With the addition of
some heat, the ice melts and the water that is released reacts with the rock to aqueously
alter the body. If enough heat is added, the parent body starts to melt and separate into
crust, mantle, and core. Illustration by Jeff Dixon.
Meteors, Meteorites, and Meteoroids  55

In addition to finding thermal metamorphism in chondrites, many car-

bonaceous chondrites (and a few other chondrites) also have been aque-
ously altered, with minerals that have had reactions with water. In some
cases, water (or a part of water, like an OH ion) actually becomes part of
the new minerals after the reaction. CI chondrites are almost entirely
made up of aqueously altered minerals containing water or OH. Some
meteorites (like some CV and CO meteorites) show evidence of once hav-
ing these types of minerals, but then experiencing further heat and
There are even some meteorites from which some fluids have been recov-
ered. Interestingly, those meteorites are ordinary chondrites, whose parent
bodies were not thought to contain much water. Taken as a whole, these
findings suggest that water may be very common on chondrite parent
bodies, and are suggestive to some that much of Earths water may have
arrived via meteorite impact.
Some processes continue to this day on meteorite parent bodies and can
be seen in the meteorites that make it to the ground. Impact melting, men-
tioned previously, occurs as some of the kinetic energy of an impactor is
converted into heat, via shocks. While this heat dissipates relatively quickly,
it can be enough in some cases to melt areas close to the impact point.
Often, meteorites with impact melt also contain unmelted fragments.
Rock that has been impacted can also experience changes, and those
changes are seen in meteorites. In the simplest case, impacts break large
rocks into smaller pieces. On parent body surfaces, this process eventually
results in the creation of a loose powder, the regolith. However, unless
regolith powder is held together somehow, it cannot survive the chain of
events that results in a meteorite. Regolith breccias are the meteorites
that result from the cementation of regolith, the process by which
impact-generated shock and pressure makes solid rock out of the pow-
dery regolith.
Impacts also create shock waves that propagate through the target. These
shock waves create very high temperatures and pressures, but only for a
very short time. As mentioned already, often some melting takes place. Even
without melting, however, shocks damage minerals inside rocks, with
results that are visible under the microscope. By conducting experiments
with shocks of known pressures, the specific effects seen in particular mete-
orites can be converted to a value for the shock pressure that meteorite
experienced. Interestingly, shocks from impacts are also thought to be re-
sponsible for cementation of regolith into regolith breccias.
In addition to these violent, sporadic events, meteorite parent bodies
also experience gentler, but more constant processes. An example of such
a process that can be seen in the meteorites is interaction with and capture
of the solar wind. Along with photons, the Sun is emitting a steady stream
of particles, mostly protons and electrons. In addition to those particles,
however, roughly 5 percent of the total are elemental nucleimostly

helium, but other elements as well. The solar wind particles can become
implanted in parent body regoliths, and many such meteorites have been
found, called gas-rich meteorites. Particles with enough momentum,
whether from the solar wind or cosmic rays from beyond the solar system,
can damage the crystalline structure of rock, as well. Because they cannot
penetrate terribly far, finding tracks from the solar wind or cosmic rays in
a sample implies that it spent time near the surface of its parent body.
Knowing or estimating the amount of damage that would accumulate
with time enables a rough age to be determined. The distribution of tracks
can even be used to show which way the sample was oriented on its parent


The Earth is constantly being hit by extraterrestrial material. Dust-sized

material burns up in the atmosphere as meteors (or shooting stars), while
larger pieces that survive passage to the Earths surface are called meteor-
ites. Meteorites are thought to come from asteroids (save those few that
originate on the Moon or Mars), while meteors seem to come both from
asteroids and from comets. Meteorites are classified based on their com-
positions, with most having abundances of the rock-forming elements
similar to those found in the Sun. Laboratory studies of meteorites and
cosmic dust are critical pieces in our understanding of the small bodies,
allowing better understanding of telescopic observations. These samples
have shown us that that some asteroids have experienced large-scale melt-
ing and volcanism, others have been affected by water and lesser amounts
of heat, and some are virtually unchanged since the origin of the solar


The Web pages for the Antarctic ANSMET expeditions can be found at this site.
Brother Astronomer: Adventures of a Vatican Scientist by Guy Consolmagno
includes an account of his experiences on an Antarctic meteorite expedition:
Web pages from Washington University with an abundance of material about
Web pages from the University of Arizona with material about meteorites: http://
Sky and Telescope magazine includes a history of scientific study of meteors at this
Web site:
NASAs Web page for the Genesis mission, including discussion of isotopic studies:
The Formation of the Solar
System and the Small Bodies

By studying the meteorite samples in our collections (as discussed in

Chapter 4), observing the small bodies through a variety of methods, and
visiting them with spacecraft, scientists have been able to craft a timeline
for the formation of the small bodies. In combination with observations
of the planets and other star systems, and of the Earth itself, we have the
outlines for the formation of the solar system as a whole. In this chapter,
we will consider how the solar system formed, and what secrets the small
bodies hold from those early times.


On the face of it, there would seem to be little prospect of successfully

describing the formation of the solar system since the distances are vast, and
the time is much, much longer than a lifetime. However, human beings have
been interested in the subject for thousands of years and have made untold
observations, which have led to numerous explanations for what they have
seen. We continue to be led by observations, which today are made by ever-
more-sophisticated instruments. Any theory of the solar systems formation
must explain the following observations, many of them centuries old:

1. There are several planets, which have different sizes.

2. All of the planets orbit the Sun in roughly the same plane.


3. The planets all have more-or-less circular orbits.

4. All of the planets orbit in the same direction.
5. Some planets are rocky, others are dominated by gas.


From observations of star-forming regions, along with theoretical model-

ing, astronomers believe that the Sun began its life in a giant molecular
cloud (GMC), a dense region of space full of dust and gas. These GMCs can
be quite large (50 light-years or more in size) and massive (up to hundreds
of solar masses), and are composed mostly of hydrogen and helium, with a
small amount of other molecules present. Such a cloud will naturally have
some areas that are slightly denser than others and are moving with slightly
different speeds. Given a disturbance like the shock wave generated by a
nearby supernova, these denser areas can begin to collapse to become
denser regions yet. A dense region like this is thought to have been the site
where the Suns formation occurred and is called the solar nebula. In gen-
eral cases, it is called a protoplanetary nebula or pre-stellar nebula. The
composition of the solar nebula was largely hydrogen and helium, with ev-
ery other element adding up to less than 2 percent of the mass of the system.
Of these elements, most of the hydrogen and some of the helium were cre-
ated at the formation of the universe (popularly known as the Big Bang),
while the rest of the helium and every other element formed inside of a star
through nuclear fusion or supernova explosions, or via radioactive decay of
elements that formed inside a star.
As a pre-stellar nebula collapses, any small initial random movement
becomes converted into rotation around a central axis. As collapse con-
tinues, gravitational and collisional interactions occur between parts of
the nebula moving in different directions. If two objects moving in differ-
ent directions collide and stick, the resulting larger object moves in a
direction that is the average of the two original directions. If the bodies
are different sizes, the post-collision direction is closer to the original
path of the larger body. This results in flattening the nebula from a
roughly spherical shape into a disk as motion differing from the average
motion of the nebula is quickly damped out. In addition, the presence of
gas causes a drag on the dust present in the nebula, which causes it to
drift inward toward the nebular center. At the center of the disk is a rap-
idly growing objectthe protosun, which will eventually become the
Sun. Examples of nebulae in this stage as imaged by the Hubble Space
Telescope can be seen in Figure 5.1, and a schematic diagram of the proc-
ess is shown in Figure 5.2. While the story of how the Sun (and stars in
general) formed is a fascinating one, we must now turn our attention
from this central attraction to the small amount of matter that would not
become part of the Sun.
Figure 5.1These Hubble Space Telescope images show disks around protostars in the
Orion Nebula. The still-forming stars at the centers of the disks shine brightly, while
the dust and gas is silhouetted against the brighter background. Images like these con-
firm the general theory of how our solar system formed, and also provide further infor-
mation to allow scientists to refine and update those theories. NASA.

Figure 5.2This cartoon shows a simplified side view of the protosolar nebula. Dust and gas
orbit the forming protosun. Collisions between the dust particles and interactions in the
gas causes an originally spheroidal cloud to collapse to a disk. Illustration by Jeff Dixon.
Catastrophic Theories
The nebular theories of solar system origin have only been generally accepted since the latter part
of the twentieth century. Competing ideas included the catastrophic theories. These theories
imagined the solar system as unusual and its formation to be a low-probability event. The most
common form of the theory began with a very close pass between the Sun and another star. During
the close pass, mass was drawn off of the Sun by the gravity of the passing star, and vice versa. This
mass was hypothesized to remain in orbit and eventually coalesce to form the planets. With the
advent of computers, astronomers were able to show that any mass drawn from the stars would
quickly fall back onto the stars after the close pass. Modifications to these catastrophic theories
inevitably were also shown to be inadequate, and with the realization that the nebular hypothesis
had no such problems, the catastrophic theories were abandoned.


The temperature of the solar nebula varied with distance from the proto-
sun, just as the temperature in the current solar system varies with distance
from the Sun. In addition, the nebula as a whole experienced cooling after
initially high temperatures. Some compounds and molecules are more sta-
ble at high temperatures than low ones, while others are more stable at low
temperatures. Experiments have been performed to determine the thermo-
dynamic characteristics of a great many minerals, and given a starting com-
position like the Sun, and a temperature and pressure (which varied at
different times and places in the solar nebula), geochemists can calculate
which minerals will form solids and which compounds will remain as gas.
The materials that form solids at the highest temperatures (1500 K or
higher) are called refractory. These materials, including aluminum oxide
and calcium titanium oxide, were the first to form as the nebula cooled,
and could potentially form everywhere in the nebula. As a result, refractory
minerals could be expected to be present on or in all of the objects in the so-
lar system, save those where later events changed or destroyed them. The
calcium-aluminum-rich inclusions (CAIs) found in carbonaceous chon-
drites and discussed in Chapter 4 are examples of refractory minerals, and
are the oldest solid material known in the solar system. As cooling contin-
ued, more solids could form, first metals like iron and nickel, followed by
magnesium-rich olivine and pyroxene around 1300 K. Again, these materi-
als could potentially form anywhere the temperature would allow it. This
set of mineralsolivine, pyroxene, and metalis largely what chondritic
meteorites are made of, consistent with other evidence that these are among
the oldest rocks in the solar system.
It is thought that the region of the solar nebula where the terrestrial plan-
ets formed was dominated by material with roughly chondritic composi-
tion. The chondrites are thus quite literally the material from which the
terrestrial planets were built. This would suggest that Venus, Earth, and
The Formation of the Solar System and the Small Bodies  61

Mars should have roughly the same composition when considering all of
their mass: core, mantle, and crust. By looking at the elements and minerals
present on the Earths surface and the few mantle rocks that have been
found, and comparing those to the chondrite meteorites, geochemists can
calculate the most likely composition of the deep interior of the Earth.
Hydrogen is not found in refractory materials. The large central mass
that would become the Sun had enough gravity to hold hydrogen gas in
place, but the dust grains forming in the inner solar system could not keep
their hydrogen. At a certain distance from the Sun, however, the tempera-
tures were cool enough to allow hydrogen-bearing compounds to form.
First, hydrated minerals form at around 500 K. These minerals contain
water or hydroxyl (OH) as part of their chemical structure, with some hav-
ing 10 percent or more water by weight. Hydrated minerals are commonly
found on the Earth today, including such well-known examples as talc, gyp-
sum, and asbestos.
Further still from the Sun, where the temperature reaches roughly 200
250 K, water ice can form, as depicted in Figure 5.3. Once we reach that dis-
tance, sometimes called the frost line or snow line, the amount of mass
that can condense from the nebula goes up immensely. Hydrated mineral
formation is limited by the abundance of magnesium and silicon, neither of
which is nearly as abundant as hydrogen and oxygen. When all of the mag-
nesium and silicon has been incorporated into mineral grains, no further
hydrated minerals can be made regardless of how much hydrogen or oxygen
is still free. Water ice formation, on the other hand, is limited by the much
greater amount of oxygen, so much more can be formed. The exact solar

Figure 5.3Because of differing temperatures, different compounds were stable depend-

ing on location in the solar nebula. Closer to the forming sun, only silicates and metals
can condense and be incorporated into the forming planetesimals. Beyond the distance
where water ice becomes stable, icy objects can form. This condensation sequence
accounts for the different compositions of different planets and the different natures of
the asteroids and comets. Illustration by Jeff Dixon.

distance of the frost line is still under debate, but is thought to be near 5 as-
tronomical units (AU), or five times the distance from the Earth to the Sun.
Unsurprisingly, this is where Jupiter, the largest planet in the solar system is
Beyond the frost line, additional important compounds condense as well.
We expect volatile materials, or those solids with low boiling temperatures
like carbon dioxide and methane, to have formed in the outer solar system
at low temperatures. In an idealized case, therefore, we would expect an
object in the far reaches of the solar system like Eris or Sedna to be com-
posed of a mixture of all of the condensable materials in the solar system:
refractory solids, chondritic silicates, water ice, and other volatile solids.
Large objects like Neptune would have these materials as well as the gravity
to be able to hold on to large amounts of hydrogen and helium gas.
This sequence of material creation at differing temperatures in the early
solar nebula is called the condensation sequence, with the process called
equilibrium condensation. The actual compositions of objects in the solar
system differ somewhat from what would be predicted by the condensation
sequence due to mixing, heterogeneity in the nebula, and random events
associated with further evolution. However, it does a good job of explaining
roughly what should be found in the solar system as a function of solar dis-
tance, shown in Figure 5.4. It can also be used to estimate the compositions
expected for extrasolar planets based on their distances from their central
stars, making some assumptions about the initial mix of gas and dust.

Figure 5.4 This graph shows the way the temperature in the solar nebula varied with dis-
tance from the Sun. As shown more qualitatively in the previous figure, different mate-
rials are stable at different distances. At the distance of the asteroid belt (about 24 AU,
only metals and silicates are stable. At larger distances, where comets were formed,
water ice, ammonia ice, and other compounds are also stable. Illustration by Jeff Dixon.
The Formation of the Solar System and the Small Bodies  63


The minerals condensing from the solar nebula were very small, only
roughly the size of dust. They were also orbiting the Sun in the presence
of gas, which slowed them down by gas drag and sent them spiraling
slowly inward, as mentioned before. When dust encountered other dust, it
was moving slowly enough that collisions were gentle and the pieces
tended to stick to one another, or accrete. The angular, jagged shapes of
condensing dust also helped stick pieces to one another. As accretions of
dust got larger, they would have a greater cross-sectional area and a
greater opportunity for further growth. As they grew larger, the gas drag
would grow less effectivethe force due to gas drag increases with the
area of a body (which is proportional to the radius squared), but the mass
of the body increases more quickly (proportionally to the radius cubed),
so the acceleration (equal to the force divided by the mass) gets smaller.
As larger accretions became less affected by gas drag, they had even more
opportunities for accreting smaller accretions that were still spiraling in
due to gas drag.
Eventually, accretions reached the size at which their gravity started to
affect the other bodies around them. This happens at a diameter of roughly
1 kilometer. At this size, they are conventionally called planetesimals by
astronomers. The first planetesimals to form had a great advantage over
smaller accretions in terms of capability for growth, and as they continued
to grow this advantage increased. Because of this, the process by which
planetesimal mass and size increased is called runaway growth. At these
sizes, gas drag can no longer change the orbits of planetesimals. Smaller
objects continued to spiral in toward the Sun, however, providing a steady
supply of objects for the planetesimals to accrete and allowing them to grow
larger still. Within a few million years, some planetesimals grew to be plane-
tary embryos hundreds of kilometers in size and would go on to form the
major planets.
Accretion rates were potentially high in the inner solar system, but as dis-
cussed already, the temperatures were sufficiently high that the amount of
accretable solids was limited. In the outermost solar system, temperatures
were low enough to allow a wide range of solids, but the accretion rates
were extremely slow, discouraging the creation of large planets. At a great
enough distance, the density of the solar nebula eventually reached a value
where accretion stalled out because the time between collisions was simply
too long. This distance is thought to have been reached beyond Neptune,
where the current transneptunian object (TNO) population represents the
original planetesimals that never formed into planets. Some objects did
manage to grow to planetary embryo size before stalling out, notably Pluto
and Eris.
As noted before, Jupiter sits at a sweet spot where silicates and water
ice are solids, but also where the accretion rate was relatively high. While

these were ideal conditions for growing a large planet, they were also opti-
mally bad conditions for Jupiters would-be neighbors. Proto-Jupiters
rapidly growing gravitational pull influenced the orbits of the planetesi-
mals nearby, increasing the eccentricity of the orbits, which increased the
average collisional speeds between those objects. Proto-Jupiter itself was
large enough to remain intact and continue accreting even with higher
encounter speeds, but the smaller objects in what is now the asteroid belt
were not massive enough to have sufficient gravity to pull themselves
back together after a large collision, and encounters on average became
corrosive rather than accretionary. As in the Kuiper belt region, some
objects (like Ceres, Pallas, and Vesta) did manage to make it to planetary
embryo-size before Jupiters growth preempted their own. And again, as
in the Kuiper belt region, the majority of the population represents the
original planetesimal population, though in this case with a composition
more representative of the inner solar system rather than the outer solar
The composition of the major planets is an average of the composi-
tions of the objects that accreted to form them. While measuring the
compositions of the deep interiors of the major planets can only be done
indirectly at best, the fact that the asteroids and comets are the building
blocks of the large planets means that they can be used to tell us about
the compositions of those larger bodies. This concept is discussed further
in Chapter 4.

Heating and Melting in the Solar Nebula
A very early heating event is required to explain the creation of chondrules, the millimeter-sized
bits of melted silicate that give chondrites their names, as discussed in Chapter 4. However, there is
ample evidence in the meteorites for additional processes driven by heat. Many chondrites show
evidence of metamorphism, the reactions between water and rock that we see evidence of in carbo-
naceous chondrites required some heat to begin, and achondrites all experienced temperatures hot
enough to melt rock. Where did this heat come from?
Today, the rocky and icy bodies of the solar system generate heat through the decay of radioac-
tive elements. After more than 4 billion years since the solar system began, only a few radioactive
elements (most notably uranium and thorium) have half-lives long enough to remain in abundan-
ces great enough to generate much heat. Early in solar system history, however, short-lived radioac-
tive elements could still be found.
Evidence from meteorites shows the supernova (or supernovae, if more than one) that began
the nebular collapse through shock waves also contributed short-lived radioactive elements. These
elements, like 26Al and 60Fe, have half-lives short enough that they rapidly disappeared, but they
gave off large amounts of heat as they did so. While we cannot find 26Al or 60Fe in meteorites today,
researchers have found the atoms they decayed into or daughter products, supporting the idea
that they were responsible for the heating.
The Formation of the Solar System and the Small Bodies  65


Roughly 10 million years after the Sun began fusing hydrogen, it is thought
to have emitted a strong solar wind, which removed the remaining gas and
much of the remaining dust from the solar system. This is called the
T-Tauri stage, after the first star observed to exhibit this wind. As smaller
objects stopped experiencing gas drag, this had the effect of significantly
slowing accretion.
As the planets reached their final sizes, there were still a number of plane-
tary embryos (as well as planetesimals) present on planet-crossing orbits.
As discussed in Chapter 3, objects in these orbits do not last long before
impacting other objects or being ejected from the solar system. It is thought
that over 90 percent of the original mass in the asteroid belt was removed
by Jupiter through close encounters, though the exact timing of this re-
moval is a matter of some debate. Planetesimals in the outer solar system
were also removed by the giant planets. In their case, however, current mod-
els suggest they remain bound to the Sun and form the population in the
Oort cloud, the region stretching nearly halfway to the nearest star where
long-period comets are thought to originate. Surprisingly, then, these fur-
thest members of our solar system are thought to have originated signifi-
cantly closer to the Sun than the Kuiper belt.
The giant planets had a great influence on the orbits of the remaining
small bodies at the end of accretion. Conversely, while a single encounter
with a planetesimal or planetary embryo had no measurable effect on giant
planet orbits, the sheer number of small bodies undergoing encounters had
a large cumulative effect. Jupiter, on average, scattered objects out of the so-
lar system to larger orbits, which had the effect of moving Jupiter itself to a
smaller orbit. Saturn, Uranus, and Neptune, on the other hand, scattered
objects into smaller orbits on average (where they tended to encounter
Jupiter) so that those objects slowly moved outward. As Neptune moved
(perhaps as much as several AU), it captured some planetesimals into res-
onances. This is thought to be the origin of the populations of the Neptune
Trojans (objects orbiting 60 degrees leading and trailing a planet) and pluti-
nos (objects with orbits like Plutos, in a 3:2 resonance with Neptune), and
perhaps the Jupiter Trojans as well.
It has also been suggested by modelers that the migrating planets them-
selves could have entered resonances with one another. As Saturn and Jupi-
ter move, for a time their orbits will be in a 2:1 resonance. This would have
a massively destabilizing effect on orbits in the rest of the solar system, and
would have resulted in a rapid depletion of the small body population from
the amounts expected early in solar system history to the amounts seen
today. The timing of this event also seems to coincide with the terminal
lunar cataclysm, an upsurge in the amount of basin formation on the
Moon (and presumably the Earth, though the evidence here has been erased
after billions of years). While neither the extent of the terminal lunar

cataclysm nor the exact dynamical models of migrating planets are univer-
sally accepted, further work will test these hypotheses.


The earliest scientific estimates of the age of the Earth, and of the solar sys-
tem, were limited by the experiences available to scientists. Geologists
studying the creation of sedimentary rocks recognized that at least hun-
dreds of millions of years or longer were necessary given the pace of current
sediment transport. Biologists also considered that similar amounts of time
were required for Darwinian evolution to create the diversity in species seen
today. Physicists of the nineteenth century, however, knew nothing about
nuclear fusion and theorized that the Sun was powered by contraction, con-
verting potential energy into heat. They calculated that the Sun could not
possibly live for longer than 20 million years, given that heat source. Simi-
larly, before the discovery of radioactivity, it was thought that the increase
in heat with depth below the Earths surface was due to energy from the
planets formation, which again implied a short time since creation.
With the discovery of nuclear fusion, it was realized that the Sun could
be billions of years old. And the earlier discovery of radioactivity and its
measurement in rocks led scientists to understand that the Earth itself was
billions of years old. These same techniques when applied to meteorites
showed that these objects are older than the oldest Earth rocks yet found
4.5 billion years old. Even those objects that have experienced melting have
been solid for practically the entirety of solar system history. In combina-
tion with the study of the textures of meteorites, further discussed in Chap-
ter 4, meteorites have been recognized as the oldest rocks in the solar
system, and the CAIs they can contain are seen as the oldest solid matter.
Often, other materials are dated using the CAIs as an initial reference time.


The asteroids, comets, and dwarf planets are all of central importance for
researchers studying the formation of the solar system and its earliest his-
tory. The most obvious reason is because they are leftover planetesimals (or
embryos in the case of Ceres, Pluto, Eris, and a few others). But there are
other reasons as well.
Their orbits give insight into the positions of the major planets. Simula-
tions and calculations that try to reproduce the current orbits of small
bodies have supported the idea that the giant planets were migrating in
early times. For instance, the existence of the Trojan asteroids of Jupiter and
Neptune are much easier to explain if those planets changed their semi-
major axes early in solar system history. The presence of asteroid collisional
The Formation of the Solar System and the Small Bodies  67

families shows that the asteroid belt was more densely populated at one
time, since the likelihood of large impacts today is very small, and the prob-
ability of creating the many large families we see is vanishingly small. With
the information that there once was more mass in the asteroid belt, models
of solar system formation can be made more accurate. Furthermore, that
information naturally leads to the question of where that mass went, again
leading to the idea that the major planets ejected the mass and have
migrated as a result.
Second, the major planets are so large and have experienced so much
processing and mixing that clues from their formation can only be indi-
rectly found after painstaking work, if at all. The fact that the small solar
system bodies are much closer to pristine objects gives us the opportunity
to study the creation of the solar system, and the Earth specifically. The
samples of small bodies in our collections provide scientists with material
that has been almost unchanged for billions of years. Telescopic observa-
tions can provide compositions for asteroids and comets, discussed further
in Chapter 7, which can then be used in conjunction with the meteorites to
map the changing composition of the protoplanetary nebula with solar dis-
tance. However, while significant insights into solar system formation have
been gained over the past decades, both meteorite evidence and telescopic
observations have limitationsmost meteorites have spent significant time
on the Earth, where they have been changed by weather and erosion in ways
that are sometimes difficult to unravel. Compositional estimates using tele-
scopic data can suffer from uncertainties in calibration and sometimes poor
data. For these reasons, many scientists support space missions to collect
fresh material from asteroidal or cometary surfaces and return it to Earth
where it can be studied. At least one such mission, discussed further in
Chapter 12, is likely to be completed in the next decade. However, given the
great diversity of material in the solar system and the cost of space missions,
it seems certain that the vast bulk of our knowledge will continue to be
gained via meteorite study and remote sensing.


Through observations of star-forming regions and what appear to be plane-

tary systems around other stars in their early stages, and via theoretical
models, astronomers have a relatively robust idea of how our solar system
formed. According to the models, a large cloud of gas and dust collapsed
into a disk, with the protosun in the center. The solids in the disk collected
into ever-larger objects, with the largest objects growing more quickly than
the smaller ones. While the largest objects eventually grew large enough to
become the major planets, some bodies stalled out at much smaller sizes.
Todays asteroids, comets, transneptunian objects (TNOs), and dwarf plan-
ets are mostly the remnants of that material, which was never collected into

planets and retains the physical properties it had when the solar system
formed. Because of the great changes that have occurred on the major plan-
ets since then, the small bodies of the solar system are the objects that best
represent the materials and conditions present in the earliest times through-
out the solar system.


University of Arizona Space Science Series: This series of books covers the
formation of the solar system in great detail. The most recent volume is Pro-
tostars and Planets V, published in 2007.


Web site for the Hubble Space Telescope. A great deal of data from the Hubble
Space Telescope has been used to study protoplanetary disks around other
stars. The site contains many of these images as well as explanations for them:
The Windows to the Universe Web site includes a set of pages about solar system
formation at various levels of detail (in Spanish as well as English), and a read-
ing list that includes books on solar system formation:
Sizes, Shapes, and
Companions of Small Bodies

In this chapter, we ask some of the most fundamental questions about the
small bodies of the solar system: How big are they? What shapes do they
have? Do they have moons? We will discuss the answers to these questions,
as well as the techniques used by astronomers and planetary scientists to
find those answers.


In our everyday lives, we measure the sizes of objects constantly. Sometimes

it is done directly with a ruler, but often we do it subconsciously, when we
look up at a tall building for instance, or note the height of those we pass
on the street. The sizes of asteroids, comets, and dwarf planets are of great
interest to scientists for a number of reasons. Knowing the size of an object
gives a measure of the strength of its gravity and also insight into its internal
structure, the temperature its interior is likely to have reached, how danger-
ous it would be if it were to strike the Earth, what compounds are likely to
be stable on its surface, and numerous other models that are critically
dependent upon size. Indeed, the definition of a planet is indirectly based
in part on the size of the object.
How do we measure the size of small bodies? The most obvious tactic
would be to take a picture or image of the object of interest and measure its


angular size. When combined with the distance to the object, the diameter in
kilometers can be calculated. This technique was used to determine the sizes
of the planets as early as the late-eighteenth century, using a device called a
micrometer to measure an objects angular size in the days before photogra-
phy or digital cameras. Other things being equal, a larger telescope enables
measurements of smaller and smaller objects. The unaided human eye has a
theoretical resolution of roughly 0.51 arcminute (an arcminute is 1/60 of
a degree), or about 0.01 degree. The Hubble Space Telescopes theoretical
resolution is 5,000 times greater than that.
The presence of the atmosphere creates a limit to the angular size that can
be measured from the ground, regardless of telescope size. Instabilities in the
atmosphere cause the appearance of objects to shimmer or twinkle, an effect
with the ungrammatical sounding name of seeing. Seeing limits measure-
ments to a few arcseconds for much of the Earth, though at astronomical
observatories specially chosen for good seeing, this number can go as low as
0.250.5 arcseconds. For objects with a large angular extent like the Moon, or
the planets, or some galaxies or nebulae, the effect of seeing is relatively
minor. For objects of very small angular extent, like stars, planetary satellites,
and small bodies, the seeing dominates measurements from the ground.
The major planets in the solar system are large enough that their sizes
can be easily measured directly, and by the 1800s their diameters were
known with only relatively small uncertainties. When measurements of
asteroids and comets were attempted, however, problems with the direct
measurement approach were found immediately. For comets, the diameter
of the coma is not fixed, but varies depending on temperature and distance
to the Sun, so it is not a useful quantity for measurement. In addition, the
coma does not have a sharp edge, making its extent unclear. The diameter
of the nucleus is a much more interesting quantity, but is difficult to meas-
ure with the coma interfering. These factors, combined with the small size
of the nucleus, have frustrated measurements of cometary sizes until rela-
tively recently.
Asteroids are by and large too small to measure directly. Ceres, the largest
object in the asteroid belt, only appears as large as a golf ball would at a dis-
tance of 1.5 km, with the next-largest objects, Pallas and Vesta, only roughly
half that size. Pluto appears about 10 times smaller from the Earth than
Ceres, with other TNOs appearing smaller still. Two advances have allowed
some measurements of the largest of small bodies using the direct tech-
nique, however: the Hubble Space Telescope (HST) allows observations to
be made without concern for seeing; and advanced telescopes have been
developed that use adaptive optics to measure and compensate for seeing
on very short timescales. The HST images of Ceres and Pluto in Figure 6.1
clearly show their disks. A small number of objects have been visited by
spacecraft, which allow sizes to be directly measured relatively easily and
very accurately. These include both asteroids and comets, with an image of
Comet Borrelly from the Deep Space 1 mission also shown in Figure 6.1.
Sizes, Shapes, and Companions of Small Bodies  71

Figure 6.1 These four images show a range of shapes and sizes seen in the small body
population. At top left is Pluto, nearly spherical and roughly 1200 km across, and its
satellite Charon, about half its size. At top right is Ceres, midway in size between Pluto
and Charon, and also nearly spherical. The Pluto, Charon, and Ceres images were all
taken using the Hubble Space Telescope. At bottom left is a model of the asteroid Kleo-
patra, made using radar data. Kleopatra is 217 km long, with the shorter axes 94 and
81 km. At bottom right is Comet 19/P Borrelly, with the image taken by the Deep Space
1 spacecraft. The comets visible area is 8 by 3 km. Kleopatra and Borrelly are distinctly
irregular in shape, Kleopatras shape in particular comparable to a dog bone. Top left,
Dr. R. Albrecht, ESA/ESO Space Telescope European Coordinating Facility; top right,
NASA; bottom left, NASA/JPL/Northwestern University; bottom right, NASA/JPL.

Objects that are large enough to show their shape are called resolved.
The majority of asteroids and TNOs are unresolved regardless of the tele-
scope that is used, appearing only as points of light. Even in this case, their
sizes can be measured or estimated by using other techniques.
The technique that may be most familiar is simply measuring the objects
brightness, a technique called photometry. Looking out over a valley at
night, we can roughly judge the relative distance between two houses by
noting how bright their lights are. Similarly, we can estimate the changing
distance to a motorcycle at night by watching its headlight change in bright-
ness. The brightness of a small body depends upon three factors: its distance
from the Sun and Earth, its size, and the darkness or lightness of the miner-
als on its surface. This last factor is called albedo. There are several different

For a given brightness, an unresolved small body can either be very reflective
Figure 6.2
and small (like the bottom object) or larger and darker (like the top object). Without in-
formation about the albedo (or reflectivity) of the object, astronomers are required to
make an estimate of the albedo and take note of their uncertainty. If an albedo measure-
ment can be made, the true size of the object can be determined. Illustration by Jeff Dixon.

kinds of albedo, which measure slightly different things, but all represent
the fraction of light falling on a surface that gets reflected; an object with an
albedo of 0.30 reflects 30 percent of the light that falls on it. The brightness,
or flux, of a body increases with albedo, and with its area (and thus its ra-
dius or diameter). A small, high-albedo object can be just as bright as a
large, low-albedo object at the same distance, if both are unresolved. This is
schematically shown in Figure 6.2. The following equation quantifies this
Flux = S albedo area/(a2 2 ) = S albedo r 2/(a2 2)
where S is a constant related to the brightness of the Sun, a is the distance
between the object and the Sun, and D is the distance between the object
and the observer.
As will be discussed further in this section, albedos for small bodies can
vary greatly. Unfortunately, they are more difficult to measure than bright-
ness. Astronomers commonly treat objects without known albedos in one
of two ways. They can assume or guess at the albedo based on other factors,
and use that value to estimate a size, or they can simply report the bright-
ness of the light reflected from the object, which is measured in magni-
tudes. Because brightness changes with distance from the Sun and Earth,
and because those distances for an object are readily calculated if its orbit is
known, any measured magnitude can be extrapolated to a particular set of
distances to allow easy comparison among many bodies, resulting in an
absolute magnitude (H). The absolute magnitude represents how bright a
small body would be at 1 AU from the Earth and 1 AU from the Sun. For
historical reasons, smaller magnitudes represent brighter objects. While two
Sizes, Shapes, and Companions of Small Bodies  73

objects with similar absolute magnitudes may have very different sizes
depending on their albedos, a large sample of bodies with similar absolute
magnitudes will on average have about the same diameters.
Because small bodies move with respect to the stars, there are occasions
when they move in front of stars and block their light. These events, called
occultations, provide another way of measuring the size of an object.
Knowing the speed of a body in its orbit (again, readily calculated for
objects with known orbits), the duration of an occultation provides a mini-
mum size. Observations of the same occultation from multiple sites can
provide additional data on the bodys size. Unfortunately, occultations are
relatively difficult to observe, since any particular event can only be seen
from a small area (much like solar eclipses are only visible from certain pla-
ces), and the best location for observing might not be known until a short
time before the event. Occultations of larger asteroids, Pluto, and Charon
have been observed in the past few decades. No comet or TNO occultations
(other than Pluto) have been observed, though opportunities for TNO
occultations in particular are eagerly sought by astronomers.
Radar observations are particularly powerful for characterizing the near-
Earth object population. The receiving dishes used in radar can be built to
be much larger than optical or infrared telescopes. In addition, the atmos-
phere interferes very little with radar. The main limitation of radar is the
amount of power required to transmit the signal. The signal loses strength
both on its way to the asteroid and on its return to the Earth. A doubling of
distance to an NEO requires a factor of 16 increase in power to achieve the
same quality of data.
Radar images are different from normal images. The radar signal is sent
from Earth at a known time with a known frequency. That frequency has a
Doppler shift upon reflection off of the object due to the objects rotation.
The time it takes the signal to arrive on Earth allows the distance to the
body (or its range) to be very accurately measured. The change in frequency
is also accurately measured, and provides the speed with which the body is
rotating. Different points on an objects surface are moving at different
speeds (for instance, one edge is moving away from the Earth, and the other
toward the Earth), and are also at slightly different distances. Therefore, so-
phisticated modeling can be required to analyze radar data.
Astronomers have also developed techniques to measure small body
albedos. As noted before, albedo measurements in conjunction with the
known distance to an object and its measured brightness allow a size to
be determined. The primary way albedos are measured is by determining
the temperature of an object. Imagine two objects, identical except that
one is white, and the other black. The same amount of sunlight falls on
both. The white object reflects the light that falls on it; the black one
absorbs it. As a result, the black object gets hotter and remains hotter
as long as they are in sunlight. This effect can be recognized by those
who like to walk barefootlighter-colored sidewalks are cooler than

asphalt-covered parking lots. The warmer an object is, the more infrared
light it emits.
The primary technique for measuring albedos is by observing objects in
the infrared spectral region, at wavelengths of 1020 mm (for comparison,
the human eye is sensitive to light from roughly 0.40.7 mm). At typical
temperatures for asteroids, comets, and TNOs, most of the light observed is
emitted black-body radiation rather than reflected solar radiation. How does
this allow albedos to be measured? By measuring the brightness of objects in
the 1020 mm region, their temperature can be calculated, and from the tem-
perature and their distance from the Sun, their albedo can be calculated. As
described previously, knowing the brightness at visible wavelengths, distance
to the object, and albedo are all that is required to calculate a size.
Why are sizes useful? As mentioned already, for individual objects the
size is an important number to include in models for their history, whether
objects have melted or not, how long they are expected to last before having
a disruptive collision, and how much damage they might do if they hit the
Earth. Knowing the proportion of larger to smaller asteroids and comets is
also useful for testing theories and models. This proportion is usually repre-
sented as a size-frequency distribution (SFD). As time goes on, objects
undergo collisions that can break them into a few pieces, or even disrupt
them into countless tiny remnants. The most destructive collisions occur
between objects of roughly the same size. Because larger objects are rarer
than small ones in the asteroid and TNO populations, a given object will
experience many more collisions with smaller objects compared to larger
ones. The larger objects have very little danger of encountering an object
large enough to disrupt them, since few of those objects exist. In addition,
the larger objects have more gravity, which helps them to better resist dis-
ruption if an impact does occur. These facts can be used to study the small
body population using the current SFD and determine what the original
SFD must have been 4.5 billion years ago. The SFDs for both main-belt aster-
oids and TNOs have been shown to be what is expected for a population
that has been experiencing mutual collisions for billions of years.

The Smallest of Small Bodies
The largest of the dwarf planets have diameters of thousands of kilometers. The largest of NEOs
have diameters that are upwards of 10 km. By extrapolating the SFD of the small body populations,
we expect millions of objects smaller than that to exist. While most of those tiny bodies are too
faint to be seen, some have been spotted while making close passes to the Earth. The smallest indi-
vidual object to be spotted in orbit around the Sun is 2006 QM111, which passed the Earth at less
than half the distance to the Moon. Based on its brightness, 2006 QM111 is thought to be
25 meters in size, which is not much larger than a human being and would easily fit in most class-
rooms. Much smaller objects commonly strike the Earth but were not spotted before entry until
late 2008, when 2008 TC3 was discovered a day before its impact.
Sizes, Shapes, and Companions of Small Bodies  75

As can be seen from Figure 6.1, asteroids and comets come in a large va-
riety of shapes, from almost round to highly irregular. As with sizes, the
most straightforward way to measure small body shapes is by direct imag-
ing. And, again, very few objects have had their shapes measured in this
way. The radar imaging described here provides an excellent way to deter-
mine the shapes of NEOs. Roughly 30 near-Earth objects have had their
shapes measured using radar.
Most shapes that are available were modeled using photometric data.
As discussed before, the brightness of an object is dependent upon its
cross-sectional area. A perfectly spherical body will always have the same
brightness as it rotates. Objects with different shapes, however, will change
their brightness as they rotate. The change in an objects brightness with
rotation is called its lightcurve. Lightcurve observations at one point in a
bodys orbit can give the relative sizes of the axes of the cross-sectional area.
Observations over several years allows different parts of an object to be seen.
When all put together, the shape of an object can be modeled.
Shapes are important for what they tell us about the interior structure of
a body. For instance, knowledge of the shape of Ceres from HST observa-
tions has allowed astronomers to deduce that body has an internal ice
ocean. The shape of an object is also a critical component for classifying it
as a dwarf planet or small solar system body. The utility of knowing the
shape of small bodies is explored in further detail in Chapter 9.


Like their larger planetary cousins, some asteroids have been found to have
satellites. Through the last decades of the twentieth century, there had been
great controversy among planetary scientists as to whether or not moons
around asteroids should be commonplace, rare, or even absent. After hints
from lightcurves suggested the presence of asteroid satellites, the first con-
firmed asteroidal satellite was found in 1993 by the Galileo spacecraft during
its flyby of Ida. This satellite, named Dactyl, led to a flurry of observations
dedicated to searching for more satellites and research designed to study the
origins of these multiple systems. Most systems have a single large body,
called the primary, with the satellite or satellites called secondaries. In the
case of Idas satellite, Ida is the primary and Dactyl is a secondary.
Some transneptunian objects are also known to have satellites. The most
famous of these systems is the Pluto-Charon system. Charon was discovered
in 1978 and is nearly the same size as Pluto, leading some to consider the
pair a double planet. In 2005, two more satellites were found around
Pluto by researchers using images from the Hubble Space Telescope, shown
in Figure 6.3. The dwarf planet Eris is also known to have a satellite, named
Dysnomia, leaving Ceres the only dwarf planet not known to have a moon.
In total, over 100 small bodies with satellites have been discovered,

Hubble Space Telescope images of the Pluto-Charon system discovered two

Figure 6.3
additional moons of Pluto in the year 2005. Multiple systems like this system are rare.
NASA/ESA/H. Weaver (JHUAPL)/A. Stern (SwRI)/HST Pluto Companion Search

including dozens of NEOs, main-belt asteroids, and TNOs. In addition to

Pluto, at least two asteroids are known to have multiple satellites.
While direct imaging can only rarely provide a measurable size or shape
for a small body, it has been an effective technique for discovering satellites.
Often, an occulting disk is used to block out the light of the central body,
which can be hundreds of times brighter than any satellites. With the cen-
tral bodys light blocked, any faint objects that may be present can be spot-
ted much more easily, and tracked through an orbit, as with the satellite of
45 Eugenia seen in Figure 6.4.
The same kinds of lightcurve observations that can be used to determine
the shape of an object are also used to determine if satellites are present. As
a primary and secondary orbit the Sun, there are times when one blocks the
other along the line of sight to the Earth. These so-called mutual events are
seen in the lightcurve as relatively sudden and sharp drops in the brightness
of the system. There are some configurations that are difficult to definitively
identify as satellites as opposed to shapesa pair of similar-sized, very
closely orbiting objects has a lightcurve that is close to that of a dumbbell-
shaped single object, for instance. Due to the relative ease of collecting
lightcurves, however, this technique has allowed many candidate binary sys-
tems to be identified.
Observations using radar have been particularly useful in finding com-
panions for small bodies. Because a satellite moves in a very different
way relative to the Earth than its primary, satellites are easily detectable
in the range-Doppler data. One of the most remarkable radar
Sizes, Shapes, and Companions of Small Bodies  77

Figure 6.4 This image shows the orbit of the satellite of Eugenia, named Petit-Prince.
Eugenia itself is at the center. Images were taken at several different times, each showing
the satellite in a slightly different place. All of those images were superimposed to create
the image above. Also superimposed is the orbit of Petit-Prince as a dashed line, and an
arrow showing its direction of motion. The cross-shaped pattern is due to stray light in
the telescope. AP Photo/European Southern Observatory (Laird Close).

Three different views of the same satellite system are shown here. The primary
Figure 6.5
is 1999 KW4, a near-Earth asteroid with a diameter of 1.2 km. The figure depicts a view
modeled using radar data, rather than being an actual image. This asteroid is rotating
very quickly, near the speed at which its own gravity could not keep it together. Most
binary NEOs are very fast rotators, which has led to theories that gravitational interac-
tions with planets or thermal forces spin up NEOs until they lose mass and satellites
form. Also notable for 1999 KW4 is the apparent ridge near its equator. Calculations
show that the rapid spin and gravitational pull of the satellite both tend to move mate-
rial toward the equator, where it accumulates. AP Photo/NASA/JPL.

observations of a satellite system is the object 1999 KW4, shown in Fig-

ure 6.5. This object is rotating quite rapidly, with a period of less than
2.5 hours. The rapid rotation plus the gravitational pull of its satellite
steadily moves material toward the equator, where a pronounced ridge
can be seen in the figure.
Naming of Satellites
For the major planets, satellite names often are related to the name of the primary. For instance,
the largest satellites of Jupiter all are named after paramours of the Roman god. Similarly, small
body satellites are often given permanent names that recall the name of the primary. Charon, the
first-discovered satellite of Pluto, is named for the ferryman who brought souls into the land of the
dead in Greek mythology. The asteroid 243 Ida is named after a mythological nymph who lived on
the very real Mt Ida in Crete. Its satellite, Dactyl, was named after other creatures that lived on that
mountain in legend. Asteroid 45 Eugenia, named in 1857 after the wife of Napoleon III, has a satel-
lite named Petit-Prince after her son, but also with a nod toward the Little Prince of literature who
lived on an asteroid. For objects without permanent names, such as 66391 1999 KW4, satellite
names are given additional numbers reflecting the name of the primary and the year of discov-
eryits satellite is S/2001 (66391) 1. While much less euphonious than Petit-Prince or Charon,
it does serve as a unique name.

All these techniques are sensitive to different sorts of satellite systems.

Direct imaging techniques are best suited to discovering and observing
objects that are well-separated, and work best on brighter (and thus typi-
cally larger) satellites. Lightcurve observations are also better suited to
larger secondaries (since they affect the shape of the lightcurve more), but
can identify close components more easily than direct imaging.


The characteristics of small body satellites are very different from those of
the satellites of the major planets. They also seem to differ between the
near-Earth object, main-belt, and TNO populations. The satellite systems
found among the NEOs generally have small, rapidly spinning primaries
with primaries having diameters less than roughly 5 km and rotation peri-
ods less than 5 hours. The primaries are also usually spherical rather than
irregular. The secondaries are most commonly less than about half the
diameter of the primary, and their separations tend to be smallonly a few
times the size of the primary itself (for instance, a 3-km-diameter primary
might have a 1-km-diameter satellite orbiting 6 km from the primarys
center, though there is variation in all of these ranges). As a whole, roughly
15 percent of NEOs are thought to be binary systems, with the fraction of
fast-spinning binary systems much larger (perhaps two-thirds).
By contrast, the percentage of binaries in the main-belt asteroid popula-
tion is much smaller among surveyed objects. Among the large main-belt
objects, only 2 percent are in multiple systems. The size differences between
primaries and secondaries in this population are much larger than that of
the NEOs, and they are typically at much larger distances. However, the
techniques used for studying main-belt asteroids cannot reach the small
Sizes, Shapes, and Companions of Small Bodies  79

sizes seen in the NEO population, making direct comparisons difficult.

Indeed, if the known NEO binary systems were moved to the distance of
the main asteroid belt, they would be very difficult to recognize as binaries.
As techniques improve and additional data are collected, astronomers will
be able to determine whether the currently known main-belt and NEO bi-
nary populations differ because they formed differently or simply because
small binary systems are different from large ones.
Binary TNOs are seen to be different from asteroids. Again, care must be
taken when comparing systems detected using quite different means, but
binary TNOs seem to be largely composed of similar-sized objects in rela-
tively distant orbits from one another. Roughly 5 percent of TNOs are
thought to be part of a binary system. Unlike the largest asteroids, few of
which have satellites, many of the largest TNOs (Pluto, Eris, 2003 EL61, and
90482 Orcus) have at least one satellite.
Because the satellites of small bodies are difficult to observe, their physi-
cal characteristics have been difficult to determine. The small amount of
available data for asteroidal satellites suggests that primaries and seconda-
ries are composed of the same material. The most detailed data for a TNO
system, that of Pluto and Charon, suggests that Charon has a higher frac-
tion of water ice than does Pluto.


By the late-twentieth century, planetary scientists had largely convinced

themselves that the small bodies did not have satellites. This was mainly
because scenarios developed to explain the formation of satellites for the
major planets were not consistent with asteroid-sized bodies. Those theo-
ries included miniature versions of the formation of the entire solar system
where a planet plays the part of the Sun, invoked to create the large satellites
of Jupiter (and not possible around smaller objects); giant impact theories
where very large collisions threw huge amounts of material into space, later
coalescing into a satellite (such impacts would utterly destroy most small
bodies); or capture via complex multi-body interactions (exceedingly
unlikely given the weak gravity of small bodies and the density of objects in
the current solar system).
The discovery of asteroidal satellites has led scientists to revisit their
assumptions. While the mechanisms for creating satellites of the major
planets are difficult to apply to the small bodies, mechanisms unique to the
asteroidal population have been identified. The first is binary creation as a
result of impact. Unlike models of the formation of the Moon, which
involve the coagulation of molten debris in Earth orbit over a long time,
these models consider the gravitational interactions of solid ejecta from an
impact into an asteroid. While the vast majority of ejecta mass is either lost
entirely or falls back onto the target object, a small amount can remain in

Figure 6.6Scientists investigating the origin of asteroidal satellites use computer simula-
tions to model large impacts, then can construct a movie of the results. The top left
shows the situation before the impact, with a small object coming in from the right to
hit a larger target. The top right portion of this figure is a still frame from the movie of
the results, and shows a large amount of ejecta, or material thrown off from the impact,
surrounding the target and moving away (in the direction of the smaller arrows), mostly
escaping. The lower left and right show the two ways we expect asteroidal satellites to be
formed. The lower left strips away most of the ejecta, showing only the central body and
one piece of ejecta that remains in orbit as a satellite (its movement shown with a curved
arrow). Finally, the bottom right focuses in on two pieces of ejecta that luckily ended up
moving in a similar direction and became bound to one another as a binary system.
These two satellite systems have very different properties: the escaping ejecta binary has a
primary and secondary of nearly the same size, like near-Earth asteroid binaries are seen
to have, while the size and orbit of the smashed target satellite is more similar to what is
seen in the main asteroid belt. Illustration by Jeff Dixon.

orbit as a satellite (sometimes called SMATS for smashed target satel-

lite). Another possibility is that two or more pieces of escaping mass could
be ejected in the same direction, allowing them to become bound to one
another into a single system (sometimes called EEBs for Escaping Ejecta
Binaries). These scenarios are depicted in Figure 6.6.
The disruption of Comet Shoemaker-Levy 9 due to the gravitational tidal
forces of Jupiter led to recognition of another possible way to create binary
systems. Simulations of small bodies passing close to planets (at reasonable
distances for NEO passes by Earth, Mars, or Venus) show that the gravity
Sizes, Shapes, and Companions of Small Bodies  81

of the terrestrial planets can tear apart fragile NEOs, which may then re-
accumulate into a binary system. However, an already-existing binary
system making a close pass to a planet is likely to be completely disrupted.
Further study is needed to see whether planetary close passes create more
binary systems than they destroy.
The combination of rapid rotation, spherical primaries, and small sizes has
suggested another mechanism for satellite creation among NEOs: spin-up and
fission. Strengthless bodies (such as asteroid rubble piles are thought to be) will
fly apart if their spin is too rapid and the centripetal acceleration at the surface
is larger than the gravitational pull. Such objects would then lose mass from
their surface as well as angular momentum, and the remaining mass would
rotate more slowly. Some of the mass that was shed could remain in orbit as
satellites. There are a few ways a NEO can be spun upvia close passes to plan-
ets and also by nongravitational forces like the YORP force, arising from the
fact that objects take time to heat up and cool down and that irregular objects
can heat unevenly (it is similar to the Yarkovsky Effect, described in Chapter
3). The rotation periods of most NEO binaries (including 1999 KW4) are close
to their rotational limit, making fission an attractive explanation.
For the known TNO binaries, the explanations used to explain asteroidal
systems do not result in systems that match observed properties. The wide
separations appear to be best matched by the multi-body capture scenarios
rejected for asteroid binaries. The objections to capture scenarios for aster-
oids are not as strong for the TNOsthe large TNO primaries have much
stronger gravity than the relatively smaller asteroidal primaries. Further-
more, while the dynamical lifetimes of asteroidal systems are relatively short,
TNO systems may be as old as the solar system itself. While the density of
the present-day asteroid belt (as well as transneptunian space) may be too
sparse to expect multi-body interactions, there was a much greater mass
present as the solar system was forming. Given the larger number of bodies
available for TNO interactions back then, it is thought that the observed
number of binary systems could have formed and survived to the present
day. The details of specific models await observations of smaller TNOs.


The previous discussion has all focused on satellites of asteroids or TNOs.

To date, no comets have been confirmed to possess a satellite. Comet Hale-
Bopp was suggested to be binary, but no firm observational evidence has
confirmed it. Comets are not expected to have satellites, since the nongra-
vitational forces on a comet (presumably acting on both primary and
secondary) would reduce the stability of any system and tend to scatter
the pieces. However, as noted earlier in the chapter, there were many good
reasons to doubt the existence of asteroidal satellites until they were
actually found.


The sizes, shapes, and presence of companions for asteroids and TNOs vary
considerably. The most straightforward way of determining these proper-
ties, direct imaging, only works for a small subset of targets at this time.
Most sizes are determined using a combination of photometry in the visible
and infrared spectral regions. Lightcurve studies have been used to deter-
mine the shapes of objects, as well as find satellites. Radar measurement, a
form of direct imaging, is a powerful tool for determining shapes and find-
ing satellites, though it is currently limited to nearby objects.
The formation of small body satellites occurred in different ways,
depending upon where the objects are found. The TNO systems we see
today are likely to have formed through capture via complex multi-body
gravitational interactions near the start of solar system history. The large
main-belt asteroids with satellites are thought to have gained those secon-
daries through collisions, either as ejecta fortuitously thrown into an orbit
that could stabilize, or as multiple pieces of ejecta thrown off in the same
direction became bound to one another. Near-Earth objects may have
formed binary systems by fission as close passes to large planets or thermal
forces increased their spin rate until they lost mass from their surface. A
large fraction of NEOs are thought to be binaries (roughly 15 percent of all
NEOs, and a much larger fraction of rapid rotators), while only a few per-
cent of main-belt asteroids and TNOs have been observed to have satellites.


Richardson, Derek, and Kevin Walsh. Binary Minor Planets. In Annual

Review of Earth and Planetary Science. Palo Alto, CA: Annual Reviews
Press, 2006.
Warner, Brian. A Practical Guide to Lightcurve Photometry and Analysis.
New York: Springer-Verlag, 2006.


This Web site has a description of the first ground-based detection of an asteroidal
satellite by the discovery team, along with movies and images: http://
This JPL radar research page contains a great amount of information on asteroid
observations, including both more technical and more popular treatments,
and many images, movies, and links:
Composition of Small Bodies

We know a great deal about the compositions of the asteroids, comets, and
dwarf planets, despite the relative scarcity of material from them available
to researchers. In this chapter we will discuss and describe the minerals
commonly found on these objects and their distribution in the solar system,
how scientists know about their presence, and how objects are classified on
the basis of their compositions.


On Earth, we are surrounded by rocks of various types: granite, basalt, mar-

ble, and sandstone, among others. These rocks are usually classified into
one of three groups, depending on their origin: igneous rocks formed when
magma or lava cooled and solidified, sedimentary rock formed from mate-
rial transported by water or wind, and metamorphic rock, which formed
from previously existing rock exposed to and changed by high heat and
Rocks are composed of groups of minerals, crystalline material some-
times made of one element, but most often compounds of several elements
in set proportions. Common minerals include gypsum, talc, and calcite
among many others. Some minerals are considered extremely valuable,
such as diamond and sapphire, while others are exceedingly common like
the quartz and magnetite that make up a large fraction of beach sand.
As discussed in Chapter 4 and 5, some elements are more commonly
found in the solar system than others. In addition, due to reasons such as


atomic size, charge, and the like, some elements are difficult to fit into crys-
tal structures and are rarely found in minerals or are only found in a few
minerals. In general, the most common minerals contain silicon, magne-
sium, oxygen, calcium, iron, and aluminum.
Quartz is the most common mineral at the Earths surface, and is com-
posed of silicon and oxygen, with a formula of SiO2. Most of the minerals
on Earth contain both silicon and oxygen, and are called silicates. The crys-
tal structure of silicates varies widely, resulting in minerals of varying prop-
erties. Other important groups of minerals include oxides (which have
oxygen, but not silicon) and carbonates (which contain carbon and oxygen
in a specific arrangement).


Two of the minerals found most commonly in Earth rocks (and in meteor-
ites) are olivine and pyroxene. Olivine is a silicate with varying amounts of
iron and magnesium. The pure magnesium version (or endmember) is
called forsterite, with a formula Mg2SiO4. The pure iron endmember is
called fayalite, and has the formula Fe2SiO4. In a crystal of olivine, small
units of Mg2SiO4 or Fe2SiO4 share oxygen atoms in making up the large
crystal. An atom of magnesium can share an oxygen atom with an atom of
iron, meaning that olivine crystals can have any mix of iron and magnesium
units. The composition of a crystal or larger sample of olivine is often
described as the percentage of forsterite and fayalite: olivine with only mag-
nesium and no iron is called Fo100Fa0 (or simply Fo100), while olivine
with half iron and half magnesium is called Fo50Fa50 (or again, simply
Fo50). Olivine is the most common mineral in the Earths mantle, and it is
also found throughout the solar system. Indeed, it has even been identified
in dust around other stars and in-between stars.
Pyroxene is a more complicated mineral than olivine. Like olivine, it is a
silicate that can contain magnesium and iron. However, its crystal structure
can also allow other elements to be incorporated, usually calcium. The pure
magnesium endmember is called enstatite, with a formula of Mg2Si2O6.
Enstatite is the major constituent of the enstatite chondrite meteorites and
the related aubrite meteorites. As with olivine, there is also an iron-rich end-
member Fe2Si2O6 (ferrosilite), and pyroxene can be found with intermediate
compositions between enstatite and ferrosilite. As mentioned, the crystal
structure of pyroxene can also accommodate some calcium. However,
because the size of a calcium ion is larger than magnesium or iron, there is
only a limited amount of calcium that can fit into a pyroxene structure.
Another major mineral on the Earth, the Moon, and in meteorites, is
feldspar. Instead of magnesium or iron, feldspar includes aluminum and
either calcium, sodium, or potassium in addition to silicon and oxygen.
Composition of Small Bodies  85

Although feldspar is commonly found in planetary and terrestrial samples,

it is difficult to detect remotely, as further discussed in another section.
The final group of silicate minerals to discuss here are the phyllosilicates.
These minerals include hydroxide (OH) or water in their crystal structure.
Phyllosilicates on the Earth form from the chemical weathering of other
minerals, and can include a number of elements such as magnesium, iron,
aluminum, potassium, and calcium, in addition to silicon and iron. The
importance of finding phyllosilicates is that their creation implies water ei-
ther currently or at some time in the past. Phyllosilicates have been seen in
meteorites, particularly the carbonaceous chondrites, and they have also
been seen on Mars and some of the asteroids.


Metals are composed of a single element, and have specific properties that
separate them from compounds. The set of atoms all share electrons, result-
ing in a greater ability to conduct electricity. Metals are relatively rare in
rocks at the Earths surface. The majority of the most common metals are
extracted from ores, like iron or copper. Those metals that occur naturally,
like gold and silver, are often considered quite valuable.
As discussed in Chapter 4, the cores of the terrestrial planets are com-
posed of iron-nickel metal, and many elements that are chemically predis-
posed to be found in metal rather than rock have been brought into those
cores as well. On the parent bodies of the chondritic meteorites, which are
undifferentiated, metal still exists at the surface. The iron meteorites are the
cores of disrupted and differentiated objects and other objects in which
metal has collected due to impact heating. When meteorites fall to Earth,
however, any iron metal present will quickly rust due to oxygen in the
Earths atmosphere.


Geochemically, organic has a different meaning than what is found at the

grocery store. An organic material is one that contains carbon, though it is
often restricted to those materials that have carbon bonded to nitrogen or
hydrogen. Much of the organic material on Earth was created by living
things, and compounds like sugar, alcohol, and amino acids are all classified
as organic.
However, organic material can also be created abiotically, without life. As
noted in Chapter 4, carbonaceous chondrite meteorites have organic mate-
rial in them, and it is also found on the surface of Saturns satellite Titan.
There is also evidence, discussed later, that it is present on comets. Organic
material is also expected on other small body surfaces like main-belt and
Trojan asteroids, and transneptunian objects. The organic compounds

found in meteorites differthe Murchison meteorite contains amino acids

including many of those found in living things. The Tagish Lake meteorite,
on the other hand, only contains simpler organic molecules. It is not clear
at this point whether the organic material in Tagish Lake is more primitive
(that is, organic material in Murchison started like that in Tagish Lake and
further chemical reactions created amino acids), or whether the material in
Murchison is more pristine (that is, Tagish Lake and Murchison had similar
kinds of organic compounds, but much of what was in Tagish Lake was
destroyed, leaving only what is seen today). It is also not clear how much
some of these samples may have been changed and contaminated by their
time on Earth, despite rapid collection and careful handling. Additional
samples from organic-rich objects are dearly desired by geochemists, who
hope to see a mission return material directly from an asteroid or comet,
which would be truly pristine material.
The formation of organic compounds in early solar system history has
been simulated in a number of experiments since the mid-twentieth cen-
tury. The most famous is the Miller-Urey experiment, in which a mixture
of water, ammonia, methane, and hydrogen was subjected to electricity sim-
ulating lightning. The resulting mixture included organic compounds
including amino acids and sugars. The organic compounds present in
meteorites and through the solar system most likely formed in similar ways
through the reactions of these common molecules. The origin of life on
Earth is a complex subject that has filled many textbooks and is still a source
of ongoing research. The role that was played by small bodies in delivering
organic material to Earth via impact is still not fully understood.


The elements most commonly found in the solar system are hydrogen, he-
lium, carbon, nitrogen, oxygen, and neon, with far more hydrogen than the
rest of the elements combined. These elements make a number of com-
pounds with one another: water (H2O), carbon dioxide and monoxide (CO2
and CO), methane (CH4), ammonia (NH3), and hydrogen cyanide (HCN),
among others. However, in the inner solar system the temperatures are too
high to allow these compounds to be incorporated into rocks (and neon and
helium are noble gases, which do not form compounds). As a result, rocks in
the inner solar system are largely composed of other elements.
At greater distances from the Sun, temperatures drop and these com-
pounds become stable. At outer solar system temperatures, water ice is as
strong as rock, while volcanoes powered by warm carbon dioxide can spew
out lava made of nitrogen (see Chapter 9).
Water ice can have several different crystal structures depending on tem-
perature and pressure. The ice we encounter on Earth is crystalline, but we
also find amorphous (or noncrystalline) ice in the outer solar system.
Composition of Small Bodies  87

Amorphous ice is a sign of quickly cooled ice, which has remained at low
temperatures since its formation.


For the vast majority of planetary studies, scientists have to gather data via
telescope or satellite rather than from the surface of the object of interest.
This is called remote sensing, as opposed to in situ (literally meaning in
the place or on-site) studies like those of the Mars Exploration Rovers or
the Huygens probe to Titan. Compositional studies typically use reflected or
emitted light from objects, although other means of getting compositional
data are discussed at the end of the chapter.


Light is peculiar. In some ways and in some circumstances it is best

described as being composed of massless particles called photons that carry
energy. In other ways and circumstances it is best described as waves with a
particular distance between crests (called the wavelength). The physics of
light is a complicated subject, and for our purposes we must content our-
selves with this abbreviated description.
The human eye can detect light with wavelengths from roughly 400700 nm,
which is called visible light. Visible light with short wavelengths (nearer
400 nm) is seen as blue; near the long wavelength end (700 nm) is seen as
red. Light can be found far beyond these bounds, however. Light with wave-
lengths longer than 700 nm is called infrared or IR (beyond red), while
ultraviolet or UV light has wavelengths shorter than 400 nm. Just as visible
light can be detected by the eye, infrared light can be detected by skin as heat.
Light can also be found beyond the infrared and ultraviolet ranges, as seen in
Figure 7.1. However, the visible, infrared, and ultraviolet ranges are the most
important for studies of asteroids, comets, and dwarf planets.
The wavelength of light and the energy it carries are connected via an
inverse relationship: E = hc/l, where E is the energy, l is the wavelength, h
is a constant (called Plancks constant), and c is the speed of light.
Astronomers can use the amount of light reflected from objects at vary-
ing wavelengths to measure their composition. This technique, called spec-
troscopy, is not limited to astronomical objects, but is also widely used on
the Earth. The changing intensity of light coming from an object at differ-
ent wavelengths is called a spectrum (plural: spectra).


Light interacts with matter in a particular way. An atom or molecule must

do one of three things to any photon that hits them: absorb it (and absorb

Figure 7.1Light is divided up into several categories depending on its wavelength. Our
eyes are only sensitive to a narrow range of wavelengths, called visible light (wave-
lengths near 0.5 mm, or 500 nm, or 0.000005 m). Shorter wavelength light carries more
energy and includes ultraviolet (UV) light, X-rays and gamma rays. Longer wavelength
light includes infrared (IR), microwave, and radio. Most compositional research for
comets, asteroids, and dwarf planets is done using IR, visible, and UV light. Illustration
by Jeff Dixon.

its energy), reflect it (changing the photons path), or transmit it (allow it

to pass through). Because of the specifics of atomic structure, only photons
of a few specific energies can interact with and be absorbed by atoms of a
particular element. Photons with different energies will not be absorbed.
Similarly, if atoms are heated by absorbing photons, they can only cool
down by giving off (or emitting) photons of certain energies. As mentioned
before, the energy of a photon is related to its wavelength, and therefore the
Composition of Small Bodies  89

The atoms in a gas can only absorb and emit light of particular wavelengths.
Figure 7.2
Seen here is the characteristic pattern of light emitted by neon atoms, familiar to us in
neon lights. The top panel shows how the light from a neon lamp would appear after
passing through a prism. The bottom panel is a graph of the amount of light as a func-
tion of wavelength, or a spectrum (in this case an emission spectrum). The wavelengths
of emission can be easily read from the spectrum. This technique is at the heart of most
compositional studies of asteroids, comets, and dwarf planets. Illustration by Jeff Dixon.

specific energies that are important for a given material mean that specific
wavelengths are important for them. A familiar example of this phenom-
enon is seen in neon lights, whose characteristic color is a result of emission
of photons that are mostly in the red part of the visible spectrum.
If we were to measure the amount of light coming from a neon lamp as a
function of wavelength, we would come up with a graph like Figure 7.2at
most wavelengths, the lamp gives off no light, but at a few wavelengths it is
bright. This particular set of lines is diagnostic of neon, and the gas in a
neon lamp in space could be identified if its spectrum were seen. Each ele-
ment has a particular combination of photon energies (and thus wave-
lengths) that it produces in a spectrum, making remote identification
possible. In fact, helium was first detected on the Sun by astronomers before
it was found on Earth. Because its spectrum was like no terrestrial material
then known, it was named for the Sun (helios in Greek).
Within a decade of the first stellar spectroscopy, the spectra of comets
were being measured for the first time. By the late 1870s, carbon com-
pounds were detected in comets. Spectroscopy has continued to be an im-
portant tool for cometary studies since those first spectral measurements.
The coma and tail of a comet are observed to have emission lines, like
the neon lamps discussed before. Some light is also solar light reflected
by the dust grains. The intensity of this reflected light changes slowly
and smoothly with wavelength, and is called the spectral continuum. The
amount of absorbed and emitted light is usually measured in comparison to
the continuum.


The fraction of light reflected from an astronomical object compared to the

light falling on it is defined as its albedo. The most commonly used albedo
values are those measured in the visible wavelength range. Although not tech-
nically compositional measures, albedo values can be useful guides to compo-
sition. As a rule of thumb, more organic-rich and carbon-rich bodies tend to
have low albedos (in the range of 510 percent, usually expressed as a decimal
range (0.050.10), with cometary nuclei generally having albedos of 0.04. Sili-
cate objects of ordinary chondrite-like compositions have albedos in the
range of 0.20.4, as do metallic objects. Ice-covered objects will have higher
albedos still, reflecting 6080 percent or more of the light that falls on them.
The albedo of an object is also an important factor in its surface temperature.
As already mentioned, photons hitting a surface will either be reflected, trans-
mitted, or absorbed. A high-albedo object will reflect most of the light that hits
it, leaving very little to be absorbed, while the opposite is true of a low-albedo
object. Light absorbed by an object will heat it up, as seen every day here on
Earth when things are left in the sunshine. Many of us also have first-hand expe-
rience with the idea that dark objects heat up more than lighter ones, as anyone
who has walked barefoot across both blacktop and sidewalks will agree!
Perhaps surprisingly, all objects emit some light. The amount and aver-
age energy of the light is strongly related to the temperature of the object.
This temperature-dependent emission, called blackbody radiation, is not
dependent upon composition, except to the degree that different composi-
tions have different albedos. By looking at the distribution of light across
many wavelengths, the temperature of an object can be measured. This is
how we know the temperature of the Sun, for instance. Asteroids, comets,
and dwarf planets all emit light in the infrared part of the spectrum, and
measurements of their temperatures are relatively common. By measuring
their brightness at visible wavelengths at the same time, however, albedos
can be determined. This is because both the visible brightness (dependent
upon the amount of sunlight reflected) and the infrared brightness (de-
pendent upon the temperature) are dependent upon the objects albedo.
This gives sufficient information to allow the albedo to be determined.
For some objects, albedo measurements are all that we have. In these cases,
scientists can estimate likely compositions based on the albedo measurement.
In some cases, the estimated composition may be incorrect, but often such
estimates are close enough. For objects of particular importance, however,
scientists are rarely satisfied with albedo measurements alone.


As an introduction, let us consider what may provide a familiar example:

light coming through the earths atmosphere. Our atmosphere is mostly
Composition of Small Bodies  91

composed of nitrogen gas, with a smaller amount of oxygen gas and a little
bit of water vapor, carbon dioxide, methane, and other gases. We know
from experience that most visible light passes through the atmosphere
unimpeded (at least on a clear day) and from this we do not expect any of
these gases to have strong absorptions in the visible region. In the UV and
IR, however, the atmosphere can be strongly absorbing. In some cases, this
has had a direct effect on humankindozone (O3) absorbs UV light at par-
ticular wavelengths, preventing much of it from reaching the surface. The
ozone holes near the Earths poles are regions of depleted ozone, which
allow increased amounts of UV to reach the surface, resulting in increased
sunburn and skin cancer. Absorptions in the IR are also importantwater,
carbon dioxide, and methane all absorb IR photons, while they are trans-
parent to visible light. Visible light from the Sun passes through the Earths
atmosphere and warms the ground, which reemits IR photons. These IR
photons are then absorbed by the water, carbon dioxide, and methane in
the Earths atmosphere, leading to an overall greater temperature (this is
the well-known greenhouse effect).
The amount of gas seen by an observer is often measured in terms of its
optical depth. This is a unitless number that represents the amount of
material required to scatter or absorb roughly one-third (actually 1/e or
1/2.718) of the light that falls on it. The mass of gas corresponding to one
optical depth is dependent upon the specific molecule and wavelength in
question. Spectroscopy of gas molecules has been important in the study of
planets for over a century. It is still the best way we have to understand the
composition of the giant planets. Small bodies are more poorly endowed
with gases than planets, but they are still an important component of com-
ets. The gases in a cometary coma are observed when they emit photons,
like the neon lamp used as a previous example. These emission spectra
have contributions from the presence of many different compounds.
For many molecules in a comet, so-called fragment species, which are
component parts of a parent molecule, are observed instead of or in addi-
tion to the original molecule. For instance, most cometary activity is driven
by the sublimation of water ice, and therefore we expect to find evidence of
water. Once in space rather than on the surface of the comet, water mole-
cules (H2O) can be broken up by high-energy photons into OH and H, or
H2 and O, all of which can be broken down even further. Alternately, the
water can lose an electron and be ionized, symbolized as H2O. All of these
breakup products have different spectral signatures, and by observing all of
them, scientists can better determine the rate at which water is leaving the
The primary fragment species that have been seen in comets are various
combinations of hydrogen, oxygen, nitrogen, and carbon. These are
thought to be derived from the breakup of water, ammonia (NH3), carbon
dioxide (CO2), carbon monoxide (CO), and various organic molecules.
Sulfur has also been detected in comets spectroscopically. However, the

parent molecule for the sulfur has not been clearly identified. Because frag-
ment species are often what is observed, and each parent molecule can pro-
duce several fragment species, any suggested parent for the sulfur will also
affect the expected amount of CH, for instance. At this point, the two most
likely suspects appear to be H2S (hydrogen sulfidethe gas that gives rotten
eggs, among other things, their distinctive smell) and OCS (carbonyl
sulfideanother unpleasant-smelling gas that is the main sulfur compound
in the Earths atmosphere). Complicated organic molecules can be broken up
in several different ways, with the fragments named by the number of carbon
atoms presentC2 fragments have two, while C3 fragments have three.
In addition to detecting different molecules, spectroscopy can be used to
measure the relative abundances of different isotopes, such as hydrogen and
deuterium. This measure, called the D/H ratio, is characteristic for each
body, as discussed further in Chapter 5. With the proper equipment, differ-
ences can also be seen between two molecules of the same composition that
are structured slightly differently. For instance, two different forms of water
(called ortho and para) can be distinguished. The relative amounts of ortho
and para water in ice is temperature-dependent, and spectroscopic observa-
tions of the ortho to para ratio in comets have been used to estimate their
formation temperaturestypically 2540 K, the temperatures found
beyond Neptune.


Spectroscopy of solid surfaces is much more complicated than spectroscopy

of gases. The absorption, scattering, and transmission of photons is de-
pendent upon the particle size and temperature of the surface, as well as the
angle of the incoming light and other factors. Typically, light can only pene-
trate into rock or mineral grains a distance roughly equal to a few times its
wavelengthonly a few micrometers for visible and infrared light, and up
to a few meters for radio observations and radar studies. Most cometary
spectra have some contribution from the coma, which greatly complicates
analysis of their surfaces. As a result, the majority of work in solid surface
spectroscopy has focused on asteroids and transneptunian objects.
Unlike free-flying molecules of gas, solids are usually organized into
structures (or crystal lattices). Atoms in a crystal lattice have some ability to
move within the lattice, but are for the most part confined to a small area.
Nevertheless, the ability to move a bit changes the absorptions seen in solids
from the narrow lines seen in gases to broader bands covering a range of
wavelengths. The wavelength range for a solid depends on composition, so
olivine with different amounts of fayalite can be distinguished, for instance.
This has led to the band center being an important parameter measured by
spectroscopists. In addition, the band depth is also an important measure.
Figure 7.3 shows schematically how these parameters are measured.
Composition of Small Bodies  93

Figure 7.3Scientists remotely measuring the compositions of small bodies typically use
parameters read from reflectance spectra. Here, several of the more important concepts
are shown using a spectrum of olivine. The band center is the vertical solid line, which
is drawn at the wavelength of minimum reflectance (roughly 1.05 mm in this spec-
trum). The continuum is an estimate of how the spectrum would look without the
absorption band and is here shown with a dashed line. The band depth is the difference
in reflectance between the continuum value and the value at the band minimum, those
points here shown with solid points. Illustration by Jeff Dixon.

By measuring the spectra of minerals and meteorites in laboratories on

Earth, spectroscopists have made great progress in tying the spectra of the
small solar system bodies to specific compositions. However, real objects
are mixtures of many minerals, not just one, some of which are spectrally
featureless. Different types of mixtures require different analytical
approaches. The simplest type of mixture is an areal or linear (sometimes
called checkerboard) mixture. In an areal mixture, the different spectral
types are separated from one anothera simplified view of the Earth is an
example. The spectrum of the Earth from far away has a contribution from
the oceans and a contribution from the continents (for the moment we will
pretend the continents have a single spectrum, ignoring different rock types
and vegetation as well as ignoring the atmosphere). A photon reflected off
of the Earth will only encounter one type of material (ocean or continent)
before coming to our telescope. Even if it is reflected more than once before
heading in our direction, it is not likely to bounce off of both ocean and
continent. In areal mixes, the spectrum that is measured is a simple combi-
nation of the different spectra in the ratio of the different areas; in the case

of the previous example, the average Earth spectrum would be 0.71 times
the oceans spectrum plus 0.29 times the continental spectrum since the
Earth is 71 percent ocean and 29 percent continent.
The second, and more complicated type of mixture is an intimate (or
nonlinear) mixture. A potentially familiar example of an intimate mixture is
typical beach sanda variety of grains of different colors. From a distance,
the different grains cannot be distinguished from one another, and a beach
appears relatively uniform. In an intimate mixture, photons typically are
reflected from more than one material before they are measured. Therefore,
there is no simple way to determine the fraction of each contributing spec-
trum. However, sophisticated theoretical models have been used to estimate
the composition of intimate mixtures from analysis of observed spectra.
Most of the minerals that make up asteroid and TNO surfaces have dis-
tinctive absorption bands in the visible and near-infrared (NIR) spectral
regions, particularly between 1,000 and 5,000 nm (15 m). This includes
almost all of the mineral types mentioned before: olivine, pyroxene, and
other silicates, organic compounds, water ice, and other ices. For this rea-
son, most compositional studies take advantage of this spectral region.
However, there has been an increasing amount of research that focuses
on the mid-infrared, usually defined as the 525 m spectral region. Sili-
cates and ices also have spectral features in this region, and several space-
craft sent to Mars have brought along instruments sensitive to the mid-IR.
This wavelength region is very sensitive to temperature, and most asteroids
have large amounts of emitted light within this wavelength range, which
can complicate analysis. In addition, the surfaces of small bodies often are
covered in particles with sizes of 525 m. This also affects the analysis
because the way light is reflected and emitted from particles changes when
their size is close to the wavelength of light. However, mid-IR observations
are of increasing interest, and the Spitzer Space Telescope has provided a
large amount of mid-IR data for asteroids and comets, which is helping to
spur further research.


The processes that occur in the regolith of small bodies discussed in Chap-
ter 8, collectively called space weathering, are thought to change their spec-
tra, potentially confusing their interpretation. Studies of lunar material
suggest that micrometeorite impacts vaporize silicate particles in the rego-
lith of airless bodies, with some of the iron in those silicates redepositing on
other regolith particles. This process changes the spectrum of the regolith
of rocky bodies from a purely silicate one to a mixture of silicate and metal,
and in extreme cases it can result in spectra that look like mostly metallic
surfaces. This change can make it more difficult to interpret the surfaces of
space-weathered objects.
Composition of Small Bodies  95

The majority of space-weathering research has focused on the Moon and

inner-belt asteroids. The processes that cause space weathering on these
objects should also be occurring with varying strength on all objects in the
solar system that lack atmospheres. The possible spectral changes are diffi-
cult to detect on outer-belt asteroids, whose spectra are relatively featureless
to begin with.
However, there is some evidence that the spectral effects of space weath-
ering can be seen in Centaurs and transneptunian objects. Experiments
simulating the exposure of ice to the solar wind and UV light from the Sun
show that the high albedo of the ices found in the outer solar system tend
to become darker, consistent with existing data. Observations of TNOs are
still few enough that firm conclusions are still in the future.


The first surveys of asteroid spectra were done in the 1970s and used visible
data only. Astronomers quickly noticed similarities between some asteroid
spectra and meteorite spectra. They separated the spectra into three groups
and somewhat optimistically gave them names reminiscent of the meteor-
ites they resembled: S for stony, C for carbonaceous, and M for metal.
All taxonomies that have followed have tried to keep similar names as the
initial taxonomy, which has allowed for some continuity but also occasion-
ally has created some confusion.
Most of the letters of the English alphabet have now been assigned to an as-
teroid spectral class. However, there are still three main groups. The S com-
plex comprises asteroids with visible-wavelength spectra shaped like an
upside-down Uabsorptions due to iron oxide at the short wavelengths and
silicates at the long wavelength create that shape. S-complex asteroids include
Ida, Eros, Apophis, and Juno. The C complex includes asteroids with relatively
flat spectra, save for the iron oxide absorption at the short wavelengths. These
asteroids tend to have low albedos, and include Ceres, Pallas, and Mathilde.
The X complex includes the M asteroids. These objects all have visible spectra
that are featureless, save for slope differences between them. The albedos of X-
complex asteroids vary, with some taxonomic schemes including the albedo
in their classifications. The C, S, and X complexes are further broken down
into classes, including (somewhat confusingly) classes named C, S, and X.
There are also some asteroid classes that dont fit in any of the main three
complexes. The two of note are the V class and the D class. The V-class
asteroids have visible spectra like Vesta, with a silicate absorption much
deeper than the S-complex objects. D-class asteroids have steep, featureless
spectra, with slopes much steeper than the X-complex asteroids. Many Tro-
jan asteroids have D-class spectra, which are also similar to the spectra of
comet nuclei and the Martian satellites Phobos and Deimos. Figure 7.4
shows example spectra of the major asteroid classes.

Figure 7.4 Asteroid spectra show great variety. Here are examples of five different impor-
tant asteroid spectral classes. The D class has low reflectance, but gets steadily brighter at
longer wavelengths. Many comets and outer-belt asteroids have D-class spectra. The M
class has a similar spectrum, though with higher reflectance. These objects are associated
with iron meteorites, though several other compositions are possible. When the absolute
brightness is taken into account, D-class asteroids have spectra much more steeply sloped
than the M asteroids. C-class asteroids are intermediate in brightness but have flat spec-
tra. The largest objects in the asteroid belt are C-class, including Ceres and Pallas, and
carbonaceous chondrite meteorites have similar spectra. The final two classes shown
both have absorptions near 1- and 2-mm, indicating the silicate minerals olivine and py-
roxene. S-class asteroids look similar to ordinary chondrites, though there are some dif-
ferences in spectral slope. The S asteroids dominate the near-Earth asteroid population
and include Ida, Gaspra, Eros, and Itokawa. Finally, the V-class asteroids are named for
Vesta. Their spectra are the same as those of the HED meteorites and have been used as
evidence that those meteorites come from Vesta. They reflect much more of the light that
falls on them than the other major classes. Illustration by Jeff Dixon.


As more and more data become available, it is increasingly clear that there
can be wide variations in composition within a single asteroid spectral class.
It is also clear that different classes can have very similar compositions.
Nevertheless, while the use of taxonomies runs the risk of oversimplifying
matters, they also serve as a convenient shorthand if understood as a tool
for guiding further work.
To move beyond taxonomy, planetary scientists have sought means of
converting reflectance spectra into quantitative measures of composition.
Composition of Small Bodies  97

Through the late-twentieth century, minerals were studied in the laboratory

to understand how their structures and compositions lead to their spectra.
These studies continue today. They have generally found that as composi-
tions change, the structure of a mineral changes, leading to shifts in band
center and band depth.


All these factors come together when using reflectance spectroscopy to

determine the composition of asteroids. The first asteroidal spectrum to be
obtained was that of Vesta, which was quickly seen to be similar to the HED
meteorites. Astronomers in the 1970s and 1980s were confident that all me-
teorite types would be easily associated with asteroids or asteroid classes.
One of the associations investigated was that of the ordinary chondrite
meteorites, the most common meteorites seen to fall to Earth (see Chapter 4),
and the S-class asteroids, the most common asteroid class in the inner aster-
oid belt. The ordinary chondrites are of interest not only because they are
common, but also because they represent material that is largely unchanged
since the formation of the planets. Determining their distribution in the
asteroid belt sheds light on conditions 4.5 billion years ago and provides
insight into how the inner planets were put together.
At the first spectra for the S-class asteroids were analyzed, they were seen
to have similar features to the ordinary chondrites: absorptions due to oli-
vine and pyroxene, and roughly the correct albedos. However, as planetary
astronomers looked at S asteroids in more detail, they realized that there
were mismatches with ordinary chondrites that were potentially quite im-
portant. The two most obvious discrepancies were that the S asteroids had
shallower absorption bands and also steeper spectral slopes. These were
exactly the changes seen in lunar samples when they suffer space weathering,
leading some to propose that the S asteroids were simply space-weathered
ordinary chondrites. Others pointed to the very different conditions present
at the Moon versus the asteroids and the lack of evidence for space weather-
ing in the meteorites, and doubted that space weathering was occurring on
asteroids to any degree. They proposed instead that the S asteroids known at
the time had all melted early in their history and werent ordinary chondritic
at all. In this view, the ordinary chondrites were only found at very small
sizes. The difference between these interpretations of S-asteroid composition
led to very different conclusions about the early solar system. The question of
what the S asteroids are made of, and where in the asteroid belt the ordinary
chondrites originate, was so central to asteroid studies that solving the
S-asteroid problem was a major goal of both the NEAR Shoemaker and
Hayabusa missions. While agreement is not unanimous, most astronomers
have become convinced that the space weathering scenario is correct.


Transneptunian objects are generally much fainter than asteroids, and as a

result it is much more difficult to obtain spectral information. A taxonomic
classification system for TNOs has only recently been developed, with four
classes: BB, BR, IR, and RR. These classes describe the visible and near-IR
spectrum using four wavelengths from 0.41.25 mm. The class names are
inspired by their spectral shapes: RR objects are the reddest (their reflectance
increases most rapidly with increasing wavelength), while the BB objects are
the bluest (their reflectance decreases with increasing wavelength, or in this
case increases least rapidly with increasing wavelength). The BR and IR classes
are intermediate in redness compared to the BB and RR classes. A limitation
of this taxonomic scheme is that it is designed for those TNOs that are fea-
tureless, and objects like Pluto and Eris are difficult to classify.
Detailed spectra have only been obtained for a few TNOs. However, spec-
tral mixing models of the sort described for asteroids have also been calculated
for TNOs. Modelers have shown that the data for the four TNO taxonomic
classes can be matched using differing amounts of organic material, water ice,
and minerals. The differences between the spectra seem to largely depend
upon the amount of organic material, though the amount and effect of space
weathering on TNO surfaces is not well known, and could play a major role.
Independently of the TNO taxonomy described here, researchers have
been incorporating the more detailed spectra into a still-forming classifica-
tion scheme based on composition. This scheme also has four rough cate-
gories, splitting TNOs into methane-rich, featureless, and water ice-rich
classes. The water ice-rich class is further divided into objects with crystal-
line water ice and amorphous water ice, indicative of more-heated and less-
heated objects.


A taxonomy scheme for comets has been much more elusive than for aster-
oids or even KBOs. Cometary nuclei are rarely observed spectroscopically
because of contributions from the coma. What nuclear spectra do exist are
consistent with D-class or C-complex asteroid spectra. Because most come-
tary spectra are emission spectra from the coma, this has been the focus of
classification attempts. Unlike asteroidal spectra, which are consistent
throughout an orbit, cometary emission spectra change significantly with so-
lar distance and temperature. As a result, the spectra themselves cannot be
used for classification. However, those spectra can be used to calculate com-
positions including the relative amounts of water and organic material on an
object. It is these compositions that have been used for classification schemes.
The most promising basis for a cometary taxonomy seems to be the
abundance of the C2 and C3 fragments relative to CN and OH. The comets
Composition of Small Bodies  99

that originate in the Kuiper belt have less of these fragments than those that
originate in the Oort cloud. This has been interpreted as evidence of differ-
ent abundances of organic material in the formation locations for those
bodies, consistent with current theories. However, in contrast to the hun-
dreds of asteroids that have been observed and classified, there are relatively
few comets for which spectra are available.


While much of what we know about small body compositions comes from
spectroscopy, scientists can get indirect information about composition by
other means. The most obvious is by studying actual samples in earth-
bound laboratories. Meteorites have provided the vast majority of detailed
information about asteroid compositions that we have, as discussed in full
in Chapter 4. Cometary samples have been returned by the Stardust space-
craft (see Chapter 12), and those samples are currently under analysis. Sam-
ple returns from asteroids are in progress, though it is not certain if they
will be successful.
Density information is available for a handful of objects. The densities of
ice, rock, and metal are quite different from one another, a fact that has
been used in some cases to make estimates of composition when only den-
sity information is available. This has been most useful when high densities
are seen, which is a sign of a metallic composition. Lower densities can be
interpreted as due to the presence of ice, or alternately as evidence of high
porosity and many fractures and cracks in the object, as discussed further in
Chapter 9.
There are other indirect measures of composition. One is observations
via radar. Metal is featureless in the visible and near-infrared spectral
regions, and because radar is sensitive to metal content, it has been the most
definitive means of identifying metal-rich objects including apparently all-
iron bodies. Another indirect means is interpretation of thermal properties.
Metal, ice, and rock heat up and cool down at different rates, a property
called thermal inertia. Measurements of thermal inertia, though difficult,
can give some insight into the composition of an objects surface.


The minerals that make up the asteroids, comets, and dwarf planets are
largely the same minerals that are found on Earth. However, they are found
in different proportions to one another, and can be found with different
compositions than seen on Earth. While a few samples of small bodies have
been retrieved by spacecraft, and meteorites are pieces of additional objects
(although the exact objects are usually unknown), the vast majority of

compositional data we have is derived from remote sensing. Spectroscopy is

the most commonly used remote sensing technique for determining com-
position, and it has been used to detect silicates, ices, organic compounds,
and others. Gas-phase spectroscopy has been used to study the coma and
tails of comets and to conclude that a variety of organic compounds and
ices exist on cometary nuclei. The vast amount of spectroscopic data avail-
able for the surfaces of asteroids has necessarily led to classification schemes
to handle the data, which can provide further insight, including suggestions
that some processes like micrometeorite impact or exposure to the solar
wind can change spectra. Classification schemes for cometary and TNO
spectra have been proposed, but the relative rarity of those data mean that
these schemes are still in their infancy.


Northern Arizona University hosts this Web site that provides further information
about the minerals found in meteorites, including chemical formulas and
A general introduction to the electromagnetic spectrum and its different regions is
on this NASA-owned Web site:
The Basics of Light contains additional information about the electromagnetic
This site maintained by the United States Geological Survey has a detailed explana-
tion of a wide variety of spectroscopic techniques at a higher technical level than
presented in this book:
The USGS provides a library of reflectance spectra for minerals at this Web site:
A large amount of remote sensing data is available for small bodies. This Web site
is an excellent resource for asteroid spectra from the Small Main Belt Asteroid
Spectral Survey (SMASS):
NASA maintains a central clearinghouse for a wide variety of remote sensing data from
both missions and earthbased research at this Web site: http://pdssbn.astro.
Surface Processes

We are famously warned by our parents and teachers not to judge a book
by its cover. But what if the book is unopenable and the cover is all we can
see? Or we have a few paragraphs, but dont know for certain which book
contains them? Planetary scientists have samples of some bodies (as seen in
Chapter 4), which have given us great insight into the formation and evolu-
tion of the asteroidal and cometary populations as a whole. We have been
able to tease out some information about small body interiors, as we will
discuss in Chapter 9. However, the bulk of our knowledge of the small
bodies and dwarf planets has been obtained by remote sensing, which is
only sensitive to the top few meters (or, depending on technique, even less!)
of their surfaces.
The surfaces of the asteroids, comets, and dwarf planets are still changing
today. They are exposed to impacts both large and very small. They are
exposed to the solar wind and harsh ultraviolet light. These changes have
implications for our understanding and interpretations of small bodies and
how what we learn from meteorites and IDPs can be best applied. In this
chapter we will look at small body surfaces and their ongoing evolution.


The small bodies do not experience rain or wind, they do not have volca-
noes that erupt lava (at least in the last several billion years). However, they
do experience impacts just as the larger planets do. These impacts come in
all sizes, from tiny ones with dust grains whose individual effects are tiny


but cumulative effects are large, all the way to catastrophic collisions that
disrupt objects and send the pieces throughout the asteroid belt, and even
eventually to impact the Sun or planets. Small impacts are more common
than large impacts, but there have been billions of years for impacts of all
sizes to occur.
The surfaces of small bodies are covered with craters as a result of these
impacts. The average collisional speed experienced by objects in the asteroid
belt is roughly 5 km/s, resulting in shockwaves upon impact. These shock-
waves transfer energy very effectively into the target material, pulverizing
solid rock into pebbles, gravel, and dust and throwing it out from the
impact site, leaving a crater behind. This pulverized material, called ejecta
(since it is ejected from the impact site), will land nearby, fly for great dis-
tances, or escape the body altogether depending on the speed of the impact,
the density of the impacted rock, and the strength of the bodys gravity.
After billions of years of impacts, the surface of a small body is expected to
be covered in the ejecta of uncounted impacts, with a powdery or gravelly
texture. If one dug through this layer, one would eventually reach solid rock
(at least on larger bodies; a further discussion of thoroughly fractured
bodies is found in Chapter 9). This layer of broken-up rock is called rego-
lith. Regolith is found on larger airless bodies like the Moon, as well as small
bodies, as demonstrated in Figure 8.1.
The shockwaves during impacts that are so effective at creating ejecta are
also effective at destroying the impacting material. Most of the meteorites
that we find on Earth originated as smaller objects that broke up in the

The surfaces of most small bodies is covered in a powdery dust called regolith.
Figure 8.1
The regolith is composed of broken-up fragments of rock created during impacts. In
addition to asteroids, it is seen on the surfaces of other airless bodies like the Moon.
This image of an astronauts footprint on the Moon shows the consistency of the rego-
lith. NASA.
Surface Processes  103

atmosphere and fell relatively slowly. The objects that are large enough to
make it to the ground intact and create craters are almost entirely vaporized
themselves in the impact. The same is thought to be true for the impactors
on the asteroids, comets, and dwarf planets; scientists do not think that
there is a large amount of mixing between the impactor and the target, and
current theories suppose that the vast majority of material in a regolith ori-
ginated on the target body itself.


At the most basic level, cratering can alter the shape of an object. Vesta, for
example, has a giant crater near its south pole, which is visible in Hubble
Space Telescope images. This large crater is thought to have been formed in
the same impact whose ejecta created the Vesta dynamical family, some of
which eventually reached Earth as the HED meteorites, as was discussed
further in Chapter 4. The asteroid Mathilde also can be seen as obviously
shaped by impacts, as all of its contours appear to be the edges of craters.
For the most part, the number of craters on a surface increases with time.
After a flurry of impacts very early in solar system history, the frequency of
impacts has evened out. Knowing the rate of impacts and counting the
number of craters on a body allows its rough age to be calculated. The exact
age is dependent upon a large number of factors that have been difficult to
quantify. For instance, we need to know how often objects of different sizes
hit a body. There are fewer large objects than small ones, but how the rela-
tive proportions of large and small bodies change for different places in
the solar system (or whether it is the same everywhere) is a matter of some
debate. The distribution of impacting bodies for a given body is called
the production function since it is the set of objects that produce craters.
While it is not clear how much the production function differs from body
to body, it is clear that it changes with time for a given bodyif nothing
else, there were more impacts at all sizes early in solar system history, with a
steady decline for billions of years.
We might imagine that the number of craters will grow with time. How-
ever, at some point enough craters have collected on a surface that any addi-
tional craters must fall on top of a crater thats already there. This situation
is called equilibrium saturation, and represents a practical limit to crater
counting. However, equilibrium saturation is reached at different times for
different size craterslarge impacts are rare enough that they only accumu-
late slowly. Small craters will degrade larger craters that they overlap, but
the large crater is usually still identifiable.
Objects like the Earth and Io are quite clearly not in equilibrium satura-
tion at any size range. These bodies have erased most of their craters
through geological activity: Io through relentless volcanic action, the Earth
mostly by erosion via wind and water. Other objects, like Mars, Venus, and

the Moon, have some areas with high crater densities, and others with lower
densities. This suggests that geological activity was active at different times
in different areas. The data we have for asteroids suggest that they are close
to crater saturation, though the images of comets suggest a complex history,
discussed further in the following section.
Some of the material thrown from craters is seen as boulders on surfaces.
With the high-resolution images available for Ida, Eros, and Itokawa, there
have been studies done of the size distribution and location of boulders. On
Eros, the distribution is consistent with their creation in the most recent
impact crater, named Shoemaker. On Itokawa, however, the boulders appear
to be too big to have been made in any of the impacts on its surface! This has
been used to argue that Itokawa is a disrupted and re-accumulated rubble
pile, and that the boulders formed when Itokawa was part of a larger body.


The collisions that make craters and create regolith are sometimes large
enough to generate ejecta blocks that are kilometers in size. Large enough
impacts can disrupt a body entirely. After a so-called catastrophic collision,
the largest remaining piece of the target is only half the size of the original
object or smaller.
These large impacts have effects that can still be seen billions of years
later. Because the ejecta and fragments of the original object had relative
speeds after the collision that were small compared to their orbital speed
around the Sun (though of course much larger than the escape speed), their
post-impact orbits remain similar. Groups of objects with similar orbits
have been identified and interpreted as collisional families or dynamical
families. These include some very large groups, including the Koronis, The-
mis, and Eos families. Dynamical families are named after their lowest-
numbered member. Family studies are a popular topic in asteroid science,
as they potentially provide a means of probing the internal structure of an
object and can also be used to determine the frequency and likelihood of
collisions early in solar system history. While large families are all quite old,
small families caused by the disruption of smaller objects are still being cre-
ated. One grouping, the Karin cluster, was the result of an impact only a
few million years agoa mere blink of an eye in comparison to the 4.5-
billion-year age of the solar system.


The formation of regolith takes time. As noted, the amount of ejecta

retained by a body depends in part on its size and the strength of its gravity.
On the smallest bodies with the weakest gravity, the vast majority of ejecta
Surface Processes  105

moves much faster than escape velocity and regolith formation is impeded
or absent. On larger objects, most ejecta is retained. As a general rule of
thumb, the speed of material in ejecta depends upon the size of the par-
ticles. Boulders move more slowly than hand-sized rocks, and both of those
types move more slowly than sand-sized particles, with dust moving most
rapidly. As a result, and in combination with the dependence of regolith
retention on ejecta speed mentioned previously, the particle sizes found in
regolith are expected to differ on objects of different size.
In studying the crater populations on the asteroids Ida, Gaspra, and Eros,
it was found that there were fewer small craters than expected given the num-
ber of larger craters. It seemed as though the smaller craters were being erased
preferentially from these asteroid surfaces. On Earth, large craters are recog-
nizable for long periods of time because smaller craters can be eroded rela-
tively quickly (or at least relatively quickly compared to the age of the
planet). Without wind and water, erosion as we know it on the Earth does
not occur on small body surfaces. However, there is still a very slow change in
the appearance of surfaces as features like craters become less sharp with time.
This is largely due to the creation of regolith and micrometeorite impacts.
These gradual, steady changes should not act to erase small craters, how-
ever. One group of scientists realized that the answer might come from
larger, infrequent impacts. Impacts create seismic waves just like earth-
quakes, which can affect the local area. According to computer modeling of
impacts, the seismic waves shake the regolith, smoothing it out to some
degree. Because the smallest craters are the shallowest craters, they are the
easiest to erase, with even relatively small impacts able to create sufficiently
large seismic waves.
Looking at the craters still present on asteroid surfaces, as well as other
clues, scientists are able to estimate the depth of the regolith on various
bodies. For Ida, Gaspra, and Eros, regolith depth is thought to be greater than
50 m in many places. For comparison, the lunar regolith is only 510 m thick.
This is likely due to the relative infrequency of collisions on the Moon com-
pared to what the asteroids experience, and the increased collisional speeds
experienced by the Moon, which could serve to greatly increase the amount of
material lost by the Moon rather than incorporated into the regolith.


High-energy ultraviolet light and x-rays from the Sun can cause particles in
a regolith to have an electric charge. Because small bodies have so little grav-
ity, the repulsion that same-charge particles have for one another in a rego-
lith can be strong enough to lift regolith grains off of the surface entirely.
This effect, called electrostatic levitation, has been observed on the Moon.
While the theory is still being studied, levitation is expected to be strongest
near the terminator, or the line on a body separating the day side from
Gold Dust
The first plans for landing on the Moon assumed the lunar surface would be bare rock.
At the time, little was known for certain about the geologic history of the Moon, and
the robotic Surveyor missions were designed to lead the way for Apollo. Among the
critical questions about the Moon was whether there had ever been volcanic activity
there, and whether the craters that dotted its surface were due to impacts or vulcanism.
The prevailing view, which has turned out to be correct, was that the craters were due
to impacts, but the low-lying dark lunar mare were in effect gigantic lava flows, and the
lighter highlands areas were of a different composition.
Thomas Gold, an American astronomer, challenged this view. He argued that the
lighter and darker areas were the same compositionally, but that meteorite impacts
generated dust that moved downhill to the lower-lying areas, where the effects of solar
wind and UV light darkened the dust. He argued that the mare could potentially have
many meters of dust, with the consistency of quicksand, and that it would be shown to
be unsafe to send astronauts. The Surveyor missions were built to deal with deep dusty
surfaces as a result, though some skeptical scientists teasingly referred to the prediction
as Gold dust.
It was subsequently shown that the lunar surface is covered by a dusty regolith, but
it has much greater packing strength than Gold feared, and the Surveyor and Apollo
missions had no issues of sinking. While Gold was proven wrong in detail about the or-
igin of the lunar mare and the depth of the regolith, he is credited with being the first
to propose that a dusty surface layer might exist on planetary bodies.

night side. In theory, the very smallest particles on asteroids could be

removed entirely, while particles larger than 1 mm or so can be moved from
one part of an object to another, settling in lower areas or shadowed areas.


Both electrostatic levitation and seismic shaking have been invoked to

explain unexpected features on Eros and Itokawa called ponds. Although
they dont contain water at all, their appearance gave them their name: flat
expanses of fine-grained regolith that seem to fill depressions on the aster-
oid surfaces. Close-up images of ponds on Eros and Itokawa are shown in
Figure 8.2. As expected from the previous discussion, the particles in the
ponds on each asteroid are differently sizedEros, with stronger gravity
than Itokawa, also has smaller regolith particles. For comparison, Itokawas
regolith is composed of particles roughly the size of gravel, while on the
Moon the regolith is more like talcum powder. On Eros, the particle size is
intermediate between those cases.
Ponds on Eros can have sizes of 30 m diameter or more, with the largest
ones found near the equator. Interestingly, the areas on Eros where ponds
are formed tend to also be the areas with the lowest gravity and also ones
Surface Processes  107

Figure 8.2 Smooth areas called ponds are found on Eros and Itokawa, the only asteroids
observed at high spatial resolution. The top panel shows these ponds inside two craters
on Eros, while the bottom shows the Muses Sea, a smooth area on Itokawa, at the
upper right. Also seen at the lower left on Itokawa is a bright-rimmed crater, one of the
most obvious craters on Itokawa. NASA/JHUAPL.

that spend a large fraction of Eros year near the terminator due to Eross
unusual orientation, with its pole nearly in the plane of its orbit. These
factors seem to support an origin via electrostatic levitation. Itokawa is
much smaller than Eros, and only has two areas that could be considered
ponds. In contrast to Eros, these areas are near the poles on Itokawa.
While analysis of ponds on Eros and Itokawa is still ongoing, it seems
likely that additional data from more asteroids will be necessary to figure
out how they formed.


An unexpected finding from early images of Marss satellite Phobos were

straight-line segments that in some cases seemed to radiate from Phoboss

Figure 8.3 Phobos, the inner satellite of Mars, has an extensive set of grooves on its sur-
face, visible here stretching horizontally. Grooves are also seen on the asteroids Ida,
Gaspra, and Eros. Their origin is not completely solved, but the leading theory is that
they represent sub-surface fractures into which regolith has partially drained. NASA/
JPL/Malin Space Science Systems.

largest crater. These segments, known as grooves, have been a matter of some
controversy. In addition to appearing on Phobos, grooves are also seen on
Ida, Gaspra, and Eros. On Phobos, grooves have been attributed to many
processes including explosive sublimation of internal ice (somewhat similar
in concept to some of the cometary and dwarf planet processes described
later) and resulting from fast-moving impact ejecta from craters on Mars.
Their presence on Gaspra, Ida, and Eros, which are not thought to have inter-
nal ice and which obviously are not orbiting a planet, require a different ex-
planation. On Gaspra and Ida, some of the grooves appear as strings of pits
roughly 100 m wide and up to a few kilometers long. The depths on Ida seem
to be a few tens of meters, though that information isnt available for Gaspra.
As discussed further in Chapter 9, many asteroids are thought to have
deep fractures that can cut through the entire object. This has led to an
alternate explanation for groovesas areas where fractures are present
beneath the regolith. In this interpretation, regolith drains partway down
the fractures, resulting in the grooves. Models suggest that the appearance
of a row of pits also fits this explanation. More recently, the Rosetta flyby of
the asteroid Steins showed an object with a line of craters that seems to
radiate from a large central crater. While analysis is still underway, this too
is consistent with the overall picture. Itokawa seems to lack grooves, con-
sistent with the idea that rather than consisting of relatively intact pieces
separated by fractures, it is an unconsolidated rubble pile composed of
smaller re-accumulated fragments with no overall structure.
Surface Processes  109


The general term for material moving downhill on the surface of a body due
to gravity is mass wasting. This is a common occurrence on the Earth, and
there is evidence it occurs on small bodies as well. Because asteroids can have
regoliths of loose material, we can imagine their very surfaces as acting like
exceedingly dry sandboxes. Any of us who have spent any time in a sandbox
know it is impossible to create sheer cliffs in dry sand. We can build a pile of
sand, but the height of that pile is related to the size of its base. Once the slope
of the pile becomes too steep, sand will cascade down the sides of the pile
until the slope is restored to a critical angle or less. This angle is called the
angle of repose, and it is important on small body surfaces.
If we consider the physics of the angle of repose, we find that it is, perhaps
surprisingly, independent of the strength of an objects gravity. Therefore, it is
relatively straightforward to calculate. It is, however, not independent of
strength. A pile of sand, which has no internal strength, is very different from
the solid rock that makes up mountains and canyons. Spacecraft looking at
the surfaces of asteroids and comets have enabled maps to be made of the dis-
tribution of slopes. Almost always, they find the slopes are less than the angle
of repose, which is consistent with the idea that their surfaces are covered
with loose regolith rather than solid rock. An intriguing exception is Comet
Wild 2, found by Stardust to have some areas that must be made of strong
material, presumably a combination of rock and ice.


To understand the regoliths of the small bodies, we must first turn to the
Moon. One of the unexpected findings from the Apollo missions was that
regolith had some properties that were very different from the rocks from
which they were derived. The spectral properties of regolith and rocks were
similar in general, but different in detail, appearing darker and redder than
the original rocks, as further discussed in Chapter 7. Geochemical analyses
showed that some regolith had accumulated relatively large amounts of
hydrogen. Further research has showed that these changes are due to expo-
sure of the regolith to micrometeorite impacts and the solar wind. The solar
wind is composed mostly of hydrogen, which sticks more easily to the
grains in regolith than to rocks as a whole, explaining the increased amount
of hydrogen in regolith. The spectral changes appear to be due to microme-
teorite impacts, which vaporize a tiny amount of material and create iron
vapor, which can then solidify and coat regolith grains, altering their spec-
tra. In the lunar regolith, samples which have more solar wind implanted
hydrogen and redder and darker spectra are called more mature, while
those that are closer to the initial rocks in their properties are called imma-
ture. Eventually, undisturbed regolith on the surface of the Moon will be

unable to hold any more hydrogen and will be fully coated in iron. This
process can take 1 billion years or longer.
However, lunar regolith is rarely completely undisturbed. The processes
that mature the lunar soil only act on its very surface, in the top few thou-
sandths of a centimeter. Below that top layer, material is largely protected
from the solar wind and micrometeorites. Larger impacts, however, will
easily penetrate much deeper. An impactor only 10 cm in diameter (roughly
hand-sized) will result in a crater that would be waist-deep for most adults,
and the formerly protected, immature material inside the crater will find
itself on the surface. This process, called gardening, is constantly occurring,
mixing more-mature regolith with immature regolith.
The same processes that are occurring on the lunar surface are thought
to occur on airless bodies throughout the solar system. As a whole, the
processes that affect and mature small body regoliths are known somewhat
confusingly as space weathering. Because there is no weather in space, this
name can be seen as an unfortunate one!
Because we do not have samples directly from asteroid, comet, or dwarf
planet surfaces, our ideas of space weathering on small bodies are still
uncertain. There has been circumstantial evidence for decades that the spec-
tra of asteroids do not match the meteorites we believe they provide, at least
in some cases. Spacecraft visits to Gaspra and Ida, and notably Eros and Ito-
kawa, showed scenes consistent with space weatheringdark regolith slid-
ing downhill revealing bright regolith, and fresh-looking areas being
brighter and less red than less-fresh areas.

Figure 8.4 This image of Eros surface from the NEAR Shoemaker spacecraft shows a view
across a crater. Bright and dark areas are shown, with darker spots at the bottom of the
crater and outside the crater. This is interpreted as due to originally brighter material
darkening with time via space weathering, and then slumping to the bottom of the
crater, leaving bright material visible on the walls of the crater. NASA/PDSSBN.
Surface Processes  111

The extent and importance of space weathering on asteroids has been a

matter of controversy for decades. It has been argued that because the
Moon is much closer to the Sun than the asteroid belt is, and the speeds
with which micrometeorite collisions occur is much faster at the Moon,
that space weathering shouldnt be as effective. An additional unknown fac-
tor is the different compositions of the Moon and asteroids, which is
thought to result in changes in how their regoliths might mature. Evidence
for space weathering is largely absent from the meteorites. Some meteorites,
called regolith breccias, have spent time at the surface of an asteroid, and
show evidence for increased amounts of hydrogen and other gases from the
solar nebula, but show no evidence for vapor-deposited iron. On the other
hand, it has been suggested that the processes that turn the powdery rego-
lith into a rock strong enough to survive the passage to Earth and become a
meteorite might also destroy much of the evidence of space weathering.
Circumstantial evidence for space weathering on asteroids using ground-
based telescopes has also been found. It has been observed that on average,
small near-Earth asteroids (NEAs) are less red and have higher albedos than
larger ones. Because small objects are broken up and destroyed by collisions
more frequently than larger ones, scientists can use the sizes of NEAs to guess
their relative ages. (It should be stressed that the meaning of age for small
bodies depends on the context. The minerals found in all small bodies of all
sizes formed billions of years ago. When discussing the age of an asteroid sur-
face, we usually are describing the time since it has been exposed to space. In
the case of large and small asteroids, we are describing the time since they
have reached their present sizes.) When putting all of this information to-
gether, the data are consistent with a scenario where younger objects have less
mature regoliths that become more mature with time, consistent with what is
seen on the Moon. Critics note, however, that this is hardly the only way to
explain the data, and that the space weathering interpretation is based on a
number of assumptions that may or may not be true.
In an effort to better understand what is going on, scientists have
attempted to re-create space weathering in their laboratories. The most
promising experiments have involved firing a laser at a simulated regolith
for very short periods of timemilliseconds or less. These very short bursts
are used to try and replicate the tiny amounts of heating and melting expe-
rienced in micrometeorite impacts. The results are comparable to what is
expected of space weathering of asteroids: reddened and darker materials.
Because the results of space weathering are most obvious for ordinary-
chondrite-like compositions, most experiments have been on those meteor-
ites and the minerals in them (like olivine and pyroxene). However, micro-
meteorites and the solar wind should also be affecting the surfaces and
regolith of all asteroids, regardless of composition. Scientists are only now
beginning to perform space weathering experiments on different meteorite
types, including the carbonaceous chondrites. Interestingly, there are hints
that the spectral effects of space weathering may be very different for these

objects, and that the destruction of organic material in outer belt asteroid
regolith may make these objects less red, in contrast to the inner-belt
Icier objects in the outer solar system can also be affected by the ultravio-
let light of the Sun and cosmic rays originating from outside the solar sys-
tem, in addition to micrometeorites and the solar wind. With long times
between collisions, astronomers expect that objects like the transneptunian
objects (TNOs) experience very little gardening. Experiments to simulate
space weathering with the theorized composition of TNOs and the condi-
tions they experience show that their spectra darken and become less red
with increased exposure.


Cometary surfaces are, in contrast to asteroidal surfaces, in a constant state

of flux. They suffer impacts and accumulate craters, but they also experi-
ence erosion and internal modification. The surface of a comet is intimately
connected to its coma, serving as the comas area of origin.
When far from the Sun, the surface and near-surface of a comet is cold
enough that any ices present are stable as solids. At these temperatures, their
surfaces experience the same sorts of processes as asteroids, though space
weathering is expected to be a very slow process at great distances from the
Sun. Nearer the Sun, however, temperatures rise high enough that ices are
no longer stable. In the vacuum of space and the low pressures present just
below cometary surfaces, liquids are also unstable and solids sublime, or
change directly into gas. This gas escapes, carrying some dust and minerals
with it and creating the coma.
It is probably obvious that the escape of gas and dust from cometary sur-
faces will also help to erase craters and other surface features. However, the
loss of ice from comets does not occur evenly across their surfaces. It has
been known for decades from telescopic observations that much of the gas
and dust comes from a small number of relatively compact areas. These
cometary jets cover only a small fraction of the comets surface area, as little
as 5 percent in some cases. The rest of the surface can be relatively inactive.
The rate at which dust is carried off with gas into the coma is quite im-
portant. If the subliming ice carries a lot of material with it from an active
area, there will be a constant supply of fresh ice seeing sunlight for the first
time. If, on the other hand, the gas mostly escapes free of dust, the ice level
will retreat further and further beneath the surface. As a result, the near-
surface regolith of a comet can become ice-free. Such an ice-free regolith,
called a lag deposit, is an effective insulator. As the comet approaches the
Sun, solar heating is unable to penetrate beyond this lag deposit to the fresh
ice below, or penetrates slowly enough that the ice remains frozen until the
comet recedes from the Sun and temperatures abate.
Surface Processes  113

Figure 8.5This series shows the creation of a lag deposit on a comet. A cross-section of
the original cometary surface is shown at left, with a mixture of ice, dust, and larger
rocks. As the comet approaches the Sun, ice sublimes, taking some dust with it (panels
moving to the right). The larger rocks, however, cannot be moved by the escaping gas,
and collect at the surface. Eventually, the rocks begin to block the ice sublimation, and
the gas production ceases from that area of the comet. The layer of rocks is called a lag
deposit. If gas production ceases on the entire cometary surface, the object is consid-
ered an extinct comet. Illustration by Jeff Dixon.

Comets from the transneptunian region or the Oort cloud making their
first visits to the inner solar system will have a large amount of ices near (or
at) their surfaces, with no lag deposit at all, and can be extremely active.
This was the case with the recent comets Hale-Bopp, Hyakutake, and
McNaught. With repeated passes by the Sun, the lag deposits grow thicker.
Comets that have been perturbed by close passes to planets in the inner
solar system or near-Earth orbits will experience higher temperatures for
longer periods of time, and may reach the point that the lag deposits are
thick enough over the entire surface to prevent any activity at all. As
touched on in Chapter 1, these objects are compositionally still comets, but
are no longer observationally any different from asteroids. Such bodies are
called extinct comets and are thought to represent somewhere between
5 and 20 percent of the near-Earth object population.
Spacecraft images of comets show surfaces that are superficially like what
is seen on asteroids, but with obvious differences. Comet Wild 2, seen by
the Stardust mission, has areas that look crater-like, but have depths that
are very unusual for their sizes compared to craters on asteroids, or other
planets and satellites. These are thought to be craters that have been modi-
fied by erosion of the surface due to gas and dust loss. Comet Borrelly, vis-
ited by Deep Space 1, found a largely smooth surface devoid of craters but
with some evidence of low cliffs from which jets were originating, as well
as hints of layering. The Deep Impact images of Comet Tempel 1 had
characteristics of Borrelly and Wild 2, with some smooth surfaces, layers,
and cliffs of roughly 20 m, as well as some craters on other parts of its


The mass loss that comets experience means that they are constantly, if
slowly, changing shape. These shape changes slowly change the rotational
properties of a comet, bringing increased sunlight to some areas, and
decreased sunlight to others. The mass loss itself also is a factor independ-
ent of the shape changes. Because the dust and gas carries momentum, the
comets orbit slowly changes, with greater changes coming with greater ac-
tivity. These non-gravitational forces are hard to predict, but must be
accounted for in precise calculations of cometary orbits.
In addition, the amount of gas loss places a rough lifetime on comets.
For instance, it is estimated that Comet Tempel 1 loses roughly 1 billion
kilograms of mass during each orbit, about the mass of the Great Pyramid
of Giza in Egypt. The mass of the entire comet is only 10,000 times greater,
meaning that after roughly 10,000 orbits the mass in the comet would be
entirely ejected. With an orbital period of less than six years, we might
expect Tempel 1 to be absent from the sky within roughly 60,000 years.
Long before then, however, Tempel 1 will likely break into smaller bits
rather than steadily lose mass to all directions, and those smaller bits will
likely evaporate more quickly. Alternately, as already described, Tempel 1
could become an extinct comet.


As larger objects, dwarf planets can have some surface processes similar to
those seen on full-fledged planets or the largest satellites, in addition to
many of the processes seen on the smaller asteroids and comets. However,
none of the dwarf planets have been visited by spacecraft, leading scientists
to make their best guesses by studying objects thought to be similar.
For Pluto, this means a comparison to Neptunes satellite Triton. Triton
and Pluto are roughly the same size and at similar distances from the Sun.
Both are thought to have similar compositions. Voyager 2 passed Triton in
1989 and returned high-resolution images of its surface. The results were
surprising, as evidence of cryovulcanismvolcanoes that erupt water or
methane rather than rockwas seen on Tritons surface. Just as volcanoes
on the Earth change its surface, cryovolcanoes on Triton serve to change its
surface by erasing craters and building mountains.
It is not clear where the energy for Tritons volcanoes originates. On the
Earth, the heat is leftover from long-lived radionuclides (further discussed
in Chapter 9). On active satellites like Io and Enceladus, volcanic activity is
thought to be powered by tidal forces and the gravitational pulls of their
central planets and other satellites. On Triton, however, the idea with the
most support is the solid state greenhouse effect. We are familiar with the
greenhouse effect, which is thought to be a major reason for climate change
Surface Processes  115

on Earth: gases such as carbon dioxide and water vapor allow visible light
to pass through and heat the ground, but are efficient at trapping reradiated
heat, which leads to an increase in temperature.
The concept of the solid state greenhouse effect is similar, except that
instead of gases in an atmosphere, it is solid nitrogen in the ground itself
that allows visible light through and traps the heat. Eventually, areas below
the surface heat up to the melting or boiling point of nitrogen and erupt.
Without close-up images for Pluto, Eris, or other transneptunian dwarf
planets, we are left to hypothesize based on Triton. Observations of crystal-
line water ice on Plutos satellite Charon have been interpreted as requiring
cryovulcanism, though there may be other explanations. Searching for evi-
dence of cryovulcanism on Pluto is one of the goals of the New Horizons
We do not have close-up spacecraft images of Ceres, nor do we have any
good ideas of what objects might be similar. There are some Hubble Space
Telescope views of Ceres surface, which have tantalizing clues, however.
There are circular areas that appear consistent with craters, and some
brighter and darker areas in general. However, the difference in brightness
across Ceres surface is quite subtle.
In anticipation of Dawns arrival at Ceres, scientists have attempted to
predict what Ceres surface will be like. Compositional studies like those

Figure 8.6Although not classified as a dwarf planet, Neptunes satellite Triton is thought
to be similar to Pluto. This image of Tritons surface, from Voyager 2, shows a large
number of dark streaks, thought to be clouds from erupting volcanoes blowing in Tri-
tons thin atmosphere. These volcanoes are thought to be due to the heating and subli-
mation of subsurface nitrogen ice, which eventually builds up enough pressure to
erupt. NASA/JPL.

described in Chapter 7 show that despite an ice-rich interior, Ceres does

not have an icy surface, and calculations of Ceres surface temperature show
that ice is only stable very close to its poles. However, there is a chance that
frost can form on shaded slopes. There is also an interesting suggestion that
Ceres surface may be covered in cracks. We know that ice is less dense than
water (we see it every time we put ice cubes in a punch bowl or glass of lem-
onade). This means that for a given mass of water, its volume increases as it
freezes. Observations of Ceres suggest that it has a deep layer of ice, dis-
cussed further in Chapter 9. At one time that layer would have been liquid,
meaning that as it has frozen its volume would have expanded. This would
have led to stresses at Ceres surface, perhaps resulting in huge faults and
cracks. An additional factor is the instability of ice on Ceres surface, as
mentioned before, which might lead to some of the same processes on
Ceres surface as are seen on cometsa lag deposit, for example.


The surfaces of the asteroids, comets, and dwarf planets have been affected
by both internal and external processes. All have suffered impacts, with the
larger objects thought to have accumulated a blanket of broken-up mate-
rial, or regolith. Craters appear on all of the small bodies for which we have
close-up images. However, sublimation of ices has modified cometary sur-
faces and erased craters on some of their surfaces. The regolith on small
bodies has redistributed itself due to shaking during impacts and simply
moving downhill, as well as perhaps via more unexpected forces like elec-
tromagnetic levitation. While dwarf planets are largely unexplored, conjec-
ture based on similar objects suggests the possibility of volcanoes and more
complex surface topography.


A technical description of seismic shaking on asteroids, including a set of explana-

tory images from NEAR Shoemaker and other missions: http://www.astro.
Professor Dave Jewitt of the University of Hawaii provides information about com-
ets and dwarf planets at this Web site, including a simplified explanation of
how lag deposits form on comets:
Small Body Interiors

While the surfaces of objects are visible to our telescopes and spacecraft, the
interiors of small bodies are of intense interest to scientists as well. Learning
about a bodys interior gives us information about the amount of heating it
has experienced and how consolidated it is. The way an object reacts to an
impact is critically dependent on its interior configuration. It is also of criti-
cal concern for those studying the best way to defend Earth against NEO
collisions. Through a variety of direct and indirect measures, we have
learned a great deal about what lies beneath the surface of asteroids, comets,
and dwarf planets, although there is a great deal that is still uncertain. In
this chapter, we look at the interior structure of small solar system bodies
and what that tells us about their formation and nature.


The interiors of objects are critically affected by the strength of the rocks
and minerals that are present. The strength of an object is most simply
described here as the amount of force per area of material that can be
applied before it breaks or deforms. There is more than one kind of
strength, depending on the direction from which the force is applied. The
two that concern us here are compressive strength, the ability to resist
crushing, and tensile strength, the ability to resist being pulled apart.
The strength of a planet-sized object increases with increasing size. This
is because with increased size comes increased gravity, which holds planets
together. The theoretical force required to break a planet is exceedingly


large, but is larger for Earth than it is for Mars, and larger for Mars than it
is for Mercury. Similarly, it is larger for an object the size of Ceres (nearly
1,000 km in diameter) than it is for something only 100 km across. At the
distances in the Kuiper belt and through the transneptunian region, water
ice is so cold it behaves like rock, so the same statement holds for those
bodies as well. If we extrapolate this behavior to smaller and smaller sizes,
we expect small objects to be much weaker than large ones.
We can directly measure the strength of small objects in the laboratory.
For those samples, the size of ones hand or smaller, strength is observed to
increase as they get smaller. This somewhat surprising result is because the
mineral crystals that make up rocks can be quite hard to crush or crack, but
the collections of minerals are much easier to separate from one another.
The presence of small cracks in rocks reduces their strength, and as rocks
get to be the size of meters or larger, they have more cracks and a greater
chance of a force applied from outside randomly aligning with some of
those cracks. If we extrapolate this behavior to larger and larger sizes, we
expect small objects to be much stronger than large ones.
Obviously, small objects cannot be both weaker and stronger than large
objects, though thats what the data suggest. The problem lies in extrapolat-
ing too far. Scientists believe that as size increases from centimeter-sized
objects, strength decreases until it reaches a minimum at some critical size,
beyond which strength again begins to increase. The two regions on either
side of the critical size have different names: the side with the largest objects
is called the gravity regime since their larger gravity is important; the side
with the smaller objects is called the strength regime. The value of the
transition size is still a matter of debate for scientists, but is thought to be
somewhere in the range of 100 m.
Usually scientists are most concerned with compressive strength, which
is most relevant to impacts. However, tensile strength is what holds objects
together when they are experiencing tidal forces while passing close to a
planet or the Sun. Models of cometary breakups under those conditions
suggest that comets have exceedingly low tensile strengths, much lower than
even snow. These results are complicated by the fact that comets may be
poorly held together to begin with (see the following discussion), but the
conclusion is certainly that comets are not strong objects by any stretch of
the imagination.
As mentioned before, even solid rock has spaces and tiny cracks. This
empty space is called porosity. A pile of rocks has porosity as well, also
called macroporosity since it is between rocks rather than within a rock.
Such a pile typically can have a porosity of 40 percent or more, so 40 per-
cent of the volume is empty. A pile of sand will have a much lower porosity.
Two objects of the same composition but different porosities will have
different densities.
There are relatively few asteroids or comets for which we have measured
densities and known compositions. For those bodies, we have good
Small Body Interiors  119

Figure 9.1 Laboratory experiments on rocks and minerals show that their strength
decreases as they get larger, due to the larger number of pre-existing cracks and flaws. On
the other hand, theoretical calculations and consideration of gravity suggests that larger
planets will be stronger than smaller ones because the force of gravity on larger objects
will make them harder to break and disperse. These two facts have led to predictions
that stony objects of roughly 100 m in diameter will be the easiest to disperse in terms of
energy needed per gram of material. Larger objects are considered to be in the so-called
gravity regime, smaller ones in the strength regime. Illustration by Jeff Dixon.

estimates of porosity. There are a greater number of objects, however, for

which we have estimated compositions that can give us estimated porosity.
We find that the largest, most massive objects have low porosities. This is
not surprising, since their gravity is large enough to crush empty areas,
much the same way pushing on a sleeping bag or air mattress will make
them denser. On the Earth, near-surface porosity can be high since there
isnt much pushing on any void space from above. Deeper in the ground,
the mass of rock pushing from above increases until void spaces cannot
support the weight above. As gravity becomes weaker on smaller bodies, the
weight of overlying rock becomes less, and void spaces can exist further
below the surface, until at last objects are small enough that voids can be
maintained through the entire object.
This can be thought of as similar to the way a house might react in a
similar situation. A house is mostly empty space that is maintained by
the floor, walls, and roof. If enough mass is piled on top of the house,
eventually the roof will give way and crush the rooms inside. This is also
true if gravity were to be magically increasedeventually, the force from
the roof itself would be too much to be supported and the house would
crush itself.
At the very smallest sizes, we expect objects with diameters of only a few
meters to tens of meters (or smaller still) to have low porosity, since they

are almost certainly single rocks much like similar-sized rocks on the Earth.
Such objects are called monoliths and are thought to have both compres-
sive and tensile strength. If a monolith were to be thoroughly cracked and
fractured, its tensile strength would disappear, but it still would have some
compressive strength. Such a body is called shattered. Finally, if a shattered
body were to be taken apart, for instance via impact, and reassembled dif-
ferently, with pieces moved and rotated relative to their original positions,
the resulting body would react very differently to stresses than a monolith
or shattered body, absorbing and localizing stress more than transmitting
it. These bodies are called rubble piles, and have significant macroporosity,
as described previously.


As discussed in Chapter 4, the small bodies accreted from pieces that were
of uniform composition depending on where they formed in the solar sys-
tem. If we look at a chondrite in detail, we find it composed of minerals of
very different densities, from grains of iron-nickel metal to crystals of oli-
vine. If we were to look at a body like Itokawa, or a parent body of chon-
dritic meteorites, we would find the same composition throughout,
regardless of depth. In the outer solar system, temperatures would be cold
enough to have accreted ice and other volatiles, and those would be found
mixed in with silicates and metal as well. This increases the range of mineral
densities found in close proximity even more. An object that has such a
mixture of minerals, and which has the same composition throughout, is
called undifferentiated.
All solar system bodies accreted some amount of radioactive materials.
As those materials decay, they give off heat. This radiogenic heat will tend
to heat up the body. All objects in the solar system began with roughly the
same concentration of radioactive elements, whether 26Al, which has almost
entirely decayed at this time, or 238U, which today is still decaying and giv-
ing off heat. Because this concentration is roughly the same, it will increase
at the same rate as an objects massa body 10 percent the mass of the
Earth will have accreted 10 percent of the Earths amount of radioactive ele-
ments, and a body 10 times larger will have accreted 10 times the amount
of Earths radioactive elements.
Objects in the solar system lose heat by radiating infrared light into
space. The amount of energy radiated increases as the surface area of a body
increaseslarger areas mean more room is exposed to space, which allows
heat to escape. How do these two effects balance out? As noted, radiogenic
heating increases with the mass of an object:

1. Heating a Mass
The density of an object is a well-known quantity,
Small Body Interiors  121

2. Density = Mass/Volume
and the volume of a sphere is also a well-known quantity:
3. Volume = (4/3)  p  (radius)3
Using these equations we can relate how the radius of an object changes as its
mass increases:
4. Mass a Density  Volume = Density  (4/3)  p  (radius)3
For our purposes, we can ignore the constant terms and be satisfied by saying:
5. Mass a Density  (radius)3
6. Heating a Mass a radius3
Increasing the radius of an object by a factor of 10 will increase the heating by
about a factor of 103, or 1,000.
What about cooling? Again as discussed here, as the surface area increases, so
does the energy radiated:
7. Cooling a Surface Area
Once more, the surface area of a sphere is known:
8. Surface Area a 4  p  (radius)2.
So we can quickly determine, again discarding the constant terms, that
9. Cooling a Surface Area a radius2.

Thus, increasing the radius of an object by a factor of 10 will allow cooling

to occur 100 times faster. But we already know that increasing by that size
will lead to 1,000 times more heating! So as objects get larger, they tend to
retain their heat more efficiently. The smallest objects can cool very rapidly
and never get very hot to begin with, while objects the size of Earth will
warm up from the decay of their radioactive materials. While the rate of
heating from the long-lived radionuclides still decaying today is not suffi-
cient to heat dwarf planets and large asteroids, short-lived radionuclides
gave off a tremendous amount of heat early in solar system history, which
could be retained by bodies the size of Vesta, Pluto, or Ceres. Typically, tem-
peratures would have increased as the short-lived nuclides gave off heat
faster than it could be radiated until some maximum temperature was
reached. As time went on, there was less remaining radioactive material, so
the rate of heating went down. Eventually, objects were able to radiate heat
faster than they were generating it, and they began to cool. The maximum
temperature that was reached is dependent upon the size of the object, the
length of time it took to accrete, and other factors that are more difficult to
determine like the amount of regolith on the surface, which tends to insu-
late objects compared to those objects without regolith. Some of the short-
lived radionuclides have half-lives so short that we expect none of the atoms
to remain. For instance, 182Hf has a half-life of 9 million years. After 500
half-lives, the amount remaining is too small to be detectable. These so-
called extinct radionuclides can still be studied because of the excess of the
stable decay products that they left behind, in this case 182W.

As noted before, objects in the asteroid belt formed as mixtures of rock

and metal, some of which survive as the parent bodies of the chondritic
meteorites discussed in Chapter 4. Smaller objects reached maximum tem-
peratures that were relatively cool. At some critical size, however, the maxi-
mum temperature became hot enough to melt first the metal and then the
rock. When this happened, the liquid metal rapidly began to sink through
the liquid rock since it is so much denser. As a result, these objects were left
with a metal core and a rocky exterior, or mantle. This molten mantle often
would experience further volcanic processes and separate further, with a
low-density crust atop an olivine-rich mantle. This process is known as
differentiation, and objects that have experienced it are fittingly called
Objects that formed where ice is stable could experience an additional
differentiation before the rock-metal separation described previously.
When the melting temperature of ice was reached, a layer of liquid water
could form. The melting temperature of ice is, as we know, much lower
than the melting temperature of metal, so the rock and metal remain
mixed with each other, at least at first. In this case, the remaining rock-
metal mixture would sink, leaving the water as the highest layer.
Depending on the duration of the heating, the amount of initial water
ice, and the size of the object, the end result could be an icy mantle over
a rocky center, a full separation of ice, rock, and metal, or the loss of
some or all of the water and a rock-metal object like those described


Density is the best single measurement with which to understand the inte-
rior structure of an object. Armed with just the density of an object and a
few guesses about its composition, we can come up with a pretty good, if
rough, idea of some of its interior properties, and knowing the relative den-
sities of materials lets us know which materials are closer to the surface or
deeper in an object. As an example, lets look at Vesta. We suspect for a
number of reasons explained elsewhere that Vesta is a differentiated object,
with an iron core and rocky mantle and crust. About how big is its core and
how much of Vestas mass is it?
To begin, we consider Vestas density (about 3,400 kg/m3) and the typical
densities of iron metal and rock (8,000 and 3,000 kg/m3, respectively).
Because Vesta is so large, well be able to ignore its porosity, something
we would have to account for in some smaller bodies. Since were consider-
ing only rock and metal, Vestas bulk density will depend only on the rela-
tive proportion of these two components, and the fraction of rock will
equal 1 the fraction of metal.
Small Body Interiors  123

10. Volume fraction of rock volume fraction of metal = 1.

11. Density of Vesta = density of rock  volume fraction of rock density of
metal  volume fraction of metal.
12. Density of Vesta = density of rock  (1 volume fraction of metal) den-
sity of metal  volume fraction of metal.
Since we know most of the numbers we need, we can pretty quickly solve for the
fraction of metal in Vesta:
13. 3,400 = 3,000  (1 volume fraction of metal) 8,000  volume fraction
of metal.
The fraction of metal works out to roughly 0.08, so 8 percent of Vestas volume
is found in the metal core. However, were interested in the size of the core, so
were not quite finished. Because volume goes as radius cubed (volume ar3), we
take the cube root of 0.08 to find the fractional radius of Vestas core and find it
to be 0.43. With Vestas radius of about 265 km, the radius of the core is thus
roughly 115 km.
What about the mass fraction? With the volume fraction of the core and know-
ing that density = mass divided by volume, we can find that the
14. Mass of core = Density of core  Volume of core
15. Mass of Vesta = Density of Vesta  Volume of Vesta
16. Mass of core/Mass of Vesta = (Density of core  Volume of core) / (Density
of Vesta  Volume of Vesta) = (Density of core / Density of Vesta)  Vol-
ume fraction of core
17. Mass of core/Mass of Vesta = 8000/3400  0.08 = 0.19.

So roughly 20 percent of Vestas mass is in its core. While we liberally

rounded off numbers in the previous equations, and even though we
ignored the fact that Vesta isnt quite spherical, the answer we determined is
within the range of core sizes scientists expect for Vesta.
A similar exercise can be performed on icy bodies, substituting the den-
sity and fraction of ice for the density and fraction of metal, and remember-
ing the rock is in the center rather than on the outside in that case! More
complicated combinations of components can also be performed, for
instance for a body with ice, rock, and metal, though the results become
more uncertain as additional ingredients are added. As mentioned already,
porosity needs to be considered for smaller objects, and as bodies become
planet-sized, the density of rock can change as depths increase, which
makes the calculations more complicated. As a quick way of understanding
objects, however, this simple calculation is a good tool.


Of course, scientists have used more complex techniques and have learned a
great deal more about Vesta than simply an estimate of its core size. We are
immensely helped by the existence of the HED meteorites and the

recognition that they come from Vesta. Among the data used to deduce that
Vesta has a core is the fact that the HED meteorites are the products of
melting and they are missing siderophile elements, or those that chemically
are more likely to be found with metal, relative to the amount of those ele-
ments found in the undifferentiated, chondritic meteorites. This means that
a metallic core was formed, taking the siderophile elements as it formed.
Radioactive elements can also be siderophiles or lithophiles, and in some
cases they decay from one type of element into another. This allows further
insight into the core formation process. For instance, Hf is a radioactive
lithophile element, while W is a siderophile. At the time of core formation
on an object, the W would be taken into the core, while Hf would stay
behind in the crust and mantle. After core formation, radiogenic 182W
would continue to be created from 182Hf, however. By studying the amount
of 182W found in the HED meteorites, geochemists have found that the dif-
ferentiation of Vesta occurred very early on, within the first 515 million
years of solar system history.
Scientists suspect more than 100 objects differentiated in the asteroid
belt, given the characteristics of the iron meteorites that have been analyzed.
These objects are difficult to find in space, however. We might expect that
with 100 differentiated objects, there should be 100 objects like Vesta pres-
ent in the asteroid belt. However, Vesta is unique in being a very large object
nearly unanimously considered to be differentiated. One implication is that
the other differentiated objects may have been victims of catastrophic colli-
sional disruption, with only fragments remaining today. Remote sensing
studies of the type described in Chapter 7 have been used to try and identify

Figure 9.2 This cartoon shows our current model of Vestas interior. From our knowl-
edge of meteorites, Vestas composition, mass and density, we can estimate its metal-
silicate ratio. Even without the kind of data available for the Earth and other planets,
we are fairly certain that Vesta has an iron-nickel core, an olivine mantle, and a basaltic
crust. Illustration by Jeff Dixon.
Small Body Interiors  125

metallic objects, which would mostly likely be the cores of disrupted bodies,
and allow us to study a planetary core in detail.


While Vesta is dry, Ceres has long been known to contain water in some
form. Using similar calculations as done in the example for Vesta, we find
Ceres to have 1520 percent water by mass. As mentioned in Chapter 6, the
shape of Ceres has been determined using images from the HST. The shape
of an object and its moment of inertia are related in an intimate way to its
interior structure, with different core densities and radii resulting in differ-
ent observable shapes.
Using the HST data, scientists found that Ceres has a rocky core with an
icy mantle 65125 km thick not far below the surface. This is consistent
with computer models of the thermal evolution of Ceres. These models
start by assuming Ceres originally had certain fractions of rock and ice, and
certain concentrations of radioactive elements, and then calculate how its
interior would change as those radioactive elements decay and give off heat.
One of the critical unknown quantities is the amount of ammonia (NH3)
that is present in Ceres. Because ammonia is a natural antifreeze, its pres-
ence makes it more difficult for water to freeze, and if enough is present on
Ceres, it is possible some liquid water still exists in its interior. The thermal

Figure 9.3This cutaway view of Ceres shows the current estimates of its interior struc-
ture: A rocky core covered with a water ice mantle, and finally a thin outer crust. These
estimates are made by knowing Ceres density and the density of the likely constituents,
as well as measures of Ceres shape from the Hubble Space Telescope. STScI.

models also suggest that Ceres could possibly be differentiated further, with
an iron core of roughly 100-km radius, similar to what is expected for Vesta.
Current research is investigating whether some other large asteroids like
Pallas and Hygiea could also have similar icy mantles.


Although they are largely unknown in many ways, scientists can make cer-
tain educated guesses about the interiors of Pluto, Eris, and other large
TNOs. As with Ceres, estimates of the ice/rock ratio for these objects can be
made (roughly 3050 percent ice on Pluto, 3580 percent on Eris). Given
the amount of radioactive materials that these objects are expected to have
accreted and their size, both Pluto and Eris, as well as other large TNOs,
are expected to have differentiated their ice from their rock, though not to
have gotten hot enough to have separated a metal core. Our knowledge of
the TNOs is limited by their distance (the large spread of possible ice/rock
ratios for Eris is because we have been unable to measure its volume as pre-
cisely as other objects) and also our limited knowledge of internal composi-
tions. Similarly to Ceres, if Pluto or Eris accreted enough ammonia, they
may still have subsurface oceans of liquid water. Other materials expected
to be present in the larger TNOs, like methanol (CH3OH), will also lower
the freezing point of water.
The large TNO Haumea has a surprisingly large density of 2,600 kg/m3,
implying a very low ice/rock ratio of only 15 percent or so. Haumea has two
satellites and several objects sharing similar orbits in a dynamical family,
believed to have formed via a large impact early in solar system history.
Because its density is so high, scientists have concluded that Haumea was
differentiated at the time of the impact, which is thought to have stripped
away much of its icy mantle, resulting in the low ice fraction we see today.
As with the asteroids, smaller TNOs cannot retain heat effectively, and
are thought to be more or less as they were early in solar system history. We
expect them to be relatively pristine mixtures of ice and rock. The densities
of smaller TNOs have been found to be as low as 500 kg/m3, roughly as
dense as wood. These smaller TNOs must thus have significant macro-
porosity, as is true of smaller asteroids.



Comets formed in the same way as asteroids, through accretion of smaller

bodies as described in Chapter 5. In the outer solar system, however, accre-
tion speeds were much slower than in the inner solar system, and the plan-
etesimals were much less consolidated than those that formed the asteroids.
Small Body Interiors  127

Because of this, it is thought that comets are held together much more ten-
uously than asteroids, and basically consist of many smaller bodies (some-
times called cometesimals) in a rubble-pile structure. A major difference
between cometary and asteroidal interiors is the presence of volatiles like
water ice and carbon dioxide ice, where the cometesimals are mixtures of
ices and rock. This led Fred Whipples initial model of cometary interiors to
be dubbed the dirty snowball model. Later models acknowledged the ma-
jority of mass was rock rather than ice, and that an icy dirtball was a more
accurate description.
Cometary disruptions are relatively common, providing much of our
knowledge of their interiors. The most famous of these breakups is prob-
ably that of Shoemaker-Levy 9, which was disrupted after it passed too close
to Jupiter, and ended up impacting that planet in 1994. It split into roughly
20 pieces. Other comet breakups include Schwassmann-Wachmann 3 in
2006, and a number of sungrazing comets. The reason for the disruptions
is not certain, and probably differs from object to object. In the case of the
sungrazers, its likely the result of passing close enough to the Sun that tidal
effects rip them apart, much like Jupiter did to Shoemaker-Levy 9. For other
comets, its possible that sublimation of volatiles plays a roleas they heat
and turn from solid to vapor, the pressure on the interior increases. Because
comets are so weak to begin with, this increased pressure could blow them

Shoemaker-Levy 9
The discovery of Comet Shoemaker-Levy 9 (usually shortened to SL9) was a boon to astronomers.
The ninth periodic comet found by the team of Carolyn and Eugene Shoemaker and David Levy,
its appearance was so unusual that it was not recognized by others who had previously imaged it.
Other comets had been seen to be temporarily captured by Jupiter, but none had been seen to have
been broken up by a close pass.
Orbital calculations suggested that SL9 was broken up less than a year before its discovery, and
more startling, that it would impact Jupiter roughly 16 months after its discovery. Practically every
professional telescope on Earth and all available spacecraft were trained on Jupiter during the pe-
riod of the impacts, including the Galileo spacecraft, then en route to Jupiter and possessing a view-
ing angle unobtainable from Eartha direct view of the impacts. However, while the impacts
themselves were on the far side of Jupiter from the vantage of Earth, the sites rotated into view
shortly thereafter and were visible from telescopes as small as 3 cm in diameter. The aftereffects
could be seen for months.
While the impacts gave great insight into the atmosphere of Jupiter, it was the pre-impact pe-
riod that was of most interest for comet researchers. By studying the distribution of SL9s pieces
and knowing its orbit around Jupiter, they estimated its original size as roughly 5 km, and found it
had only a tiny amount of strengthless than that of snow. It was also determined that Jupiter is
likely the victim of a cometary impact every few decades, though one the size of SL9 is much rarer,
perhaps once in a thousand years.
Magnetic Fields of Small Bodies
Scientists can gain insight into planetary interiors by studying the magnetic fields they generate.
The origin of planetary magnetic fields, and why they have the characteristics they do, is still only
poorly understood. However, it is generally agreed that for a planet to have a magnetic field, it
needs a layer of conducting fluid. In the Earth, this is the molten outer core, composed of iron.
Given this, we dont expect much in the way of magnetism on the small solar system bodies.
However, there are possible exceptions. If Ceres has an internal liquid water ocean, it could con-
ceivably generate a magnetic field. The same is potentially true of other larger objects, like Pluto,
although its close cousin Neptunes satellite Triton does not appear to have a strong magnetic field.
Vesta is another possibility, and some scientists have recently interpreted color variations on its
surface as evidence supporting the presence of a magnetic field on that object. Data from the HED
meteorites, originating on Vesta, have been inconclusive.
However, we find that other meteorites are magnetic. Disentangling the history of this magne-
tism, which diminishes with shock and impact, is difficult, but it is thought that these meteorites
were subjected to strong magnetic fields early in solar system history, which froze in a magnetic
field, just like everyday objects can become magnetized by close exposure to a magnet.


The interiors of asteroids, comets, TNOs, and dwarf planets vary dramati-
cally in strength. Larger objects have been able to retain radiogenic heat
more effectively, and have experienced differentiation as a result, where-
upon melting metal, rock, and ice can separate into different layers, with
the densest material sinking to form a core and a less dense mantle atop it.
Smaller objects were unable to retain heat and stayed undifferentiated. As
time went on, impacts fractured smaller bodies, with some impacts large
enough to disrupt the target object, which could then reform as a jumbled
mass known as a rubble pile. At small enough sizes (perhaps only tens of
meters in size), objects cannot reform after a disruptive impact since their
gravity is too low, and any objects of that size are thought to be single,
coherent rocks. Conversely, at large enough sizes, disruptive impacts are so
rare that most bodies have never experienced them, and so objects like
Vesta, Ceres, Pluto, and Eris (among others) are also thought to be coherent
bodies. While direct evidence about small body interiors only comes from
some meteorites at present, we have gained a great deal of knowledge from
a starting point of relatively simple estimates of composition in conjunction
with measured densities.


The breakup of Comet Shoemaker-Levy 9 led to a large amount of material being

posted on the then-new World Wide Web. A pre-impact summary of what was
Small Body Interiors  129

expected can be found at this Web site. The main page (http://www2.jpl.nasa.
gov/sl9/sl9.html) carries updates from a wide variety of sources. Other comet
impacts, these into the Sun, are described on
y2000/ast10feb_1.htm, along with images of them from the SOHO spacecraft:
This site has HST images and descriptions of the recent cometary breakup of
Comet Schwassmann-Wachmann 3:
An excellent set of online pages about asteroidal and planetary interiors and differ-
entiation are at this Web site by the American Museum of Natural History:
Astronomy magazines Web site has a popular-level description of the interior of
The Dawn mission homepage has a set of links about Vesta and its interior: http://
Small Body Atmospheres

The small bodies of the solar system are typically thought of as barren rocks
hurtling in space. However, while clouds and rain never darken the skies of
asteroids, Pluto and other larger TNOs have atmospheres that change and
winds that blow. The comae of comets have elements of planetary atmos-
pheres, though with unique aspects. In this chapter, we will look briefly at
atmospheres in general before considering the types of atmospheres that are
found (or in some cases may possibly be present, if undetected) on the aster-
oids, comets, and dwarf planets.


We are all familiar with the Earths atmosphere, the part of the planet
extending from its solid surface to the edge of space. The atmosphere con-
tains the oxygen that we breathe, hosts the weather systems that bring us
rain and snow (in amounts from life-sustaining to life-threatening), and
even serves as a form of protection against ultraviolet light and some aster-
oidal and cometary impacts. Most of the major planets have substantial
atmospheres. Some of the small bodies also have atmospheres, although
their atmospheres are much thinner than our own and have very different
compositions. In this chapter, we will discuss the atmospheres that have
been found on comets, asteroids, and dwarf planets, as well as those sus-
pected and those that appear to be absent.
By and large, the more massive a body is, the larger an atmosphere it will
have. As an object gets more massive, its gravity increases and it can hold


on to gases with smaller molecular weights. In addition, the size of an

objects atmosphere is related in a complicated manner to its surface tem-
perature and its distance from the Sun. As temperature increases, a greater
number of compounds are found in the gaseous state, enabling a larger
atmosphere to be possible. However, at these higher temperatures, on aver-
age the molecules in the gas become more energetic and move more
quickly. Thus, they are more likely to be moving too quickly to be retained
by the objects gravity. The solar distance not only controls the surface tem-
perature but also the compounds available to the body in the first place via
the condensation sequence, as was described in more detail in Chapter 5.
The outer bound of a planetary atmosphere is called the exobase. Above
the exobase, a typical atmospheric molecule is moving at an average speed
that is greater than the escape velocity, and so it can escape into space, while
below it the molecule is bound to the planet. For most major planets, the
exobase is far above the surfacefor the Earth, for instance, it is at roughly
500 km, while on Venus and Mars it is over 200 km. For some of the objects
discussed in this chapter, the exobase is at the objects surface, meaning any
atmosphere we see on those bodies requires constant replenishment since it
is constantly escaping.


When it was first discovered, and it was thought to be Earth-sized or larger,

there was a general expectation that Pluto would have an atmosphere. As
estimates of its size began to approach modern measured values, however,
many began to suspect Pluto to be airless.
The first suggestions that Pluto might indeed have an atmosphere were
based on spectroscopic measurements in the 1980s that found evidence for
methane in the gaseous phase in addition to solid methane. Thermodynamic
arguments were made in light of those measurements, showing that for
methane to be present as the spectroscopic results suggested, a second gas,
probably nitrogen, was required to make up the bulk of Plutos atmosphere.
This prediction for an atmosphere on Pluto was confirmed during an oc-
cultation in 1988. In Chapter 6, we discussed how occultations are used to
determine the size and shape of airless bodies. For objects with an atmos-
phere, the star being occulted is dimmed by the atmosphere before disappear-
ing behind the solid body. The details of this dimming provide information
about the nature of the atmosphere. For Pluto, it was found that the atmos-
phere was roughly 700,000 times thinner than the Earths atmosphere. The
occultation also provided a measurement of the atomic mass of the molecules
in Plutos atmosphere, confirming the prevalence of nitrogen.
Plutos eccentric orbit means that its average surface temperature varies con-
siderably. This has led to speculation that it may get cold enough during the plu-
tonian year for the atmosphere to freeze out, leaving the surface coated with
Small Body Atmospheres  133

frost and ice. During the warmer parts of the plutonian year, the ice would sub-
lime, creating the atmosphere again. Pluto reached perihelion in 1989, and it
was thought that atmospheric freeze-out might occur quite rapidly.
However, other occultations of stars by Pluto occurred in 2002, and the
results surprised astronomers. Instead of a diminished or absent atmos-
phere, they showed the atmosphere of Pluto had gotten twice as thick as it
was in 1988. This has now been attributed to the changing seasons on Pluto.
Shortly after perihelion on Pluto, its northern hemisphere begins autumn
and its southern hemisphere begins spring. The south pole of Pluto, in
darkness for well more than 100 Earth years, begins to heat up and volatiles
that had frozen out on the pole start to enter the atmosphere; as the north
pole cools down, it will begin to accumulate volatiles.
However, given our current knowledge about Pluto and accounting for
this seasonal effect, scientists now arent certain whether freeze-out happens
or not, though according to some models the plutonian atmosphere may
begin to freeze out around the time of New Horizons arrival in 2015.
Charon, similar to Pluto in many ways, seems to lack an atmosphere. This
is likely because its mass, and thus gravity, is just a bit too small to retain
even a tenuous atmosphere.
Even though Pluto has retained its atmosphere over the history of the so-
lar system, molecules are still escaping from above its exopause. It has been
estimated that 33,000 kg of nitrogen escapes from Pluto every second, or
the equivalent of up to several of the Egyptian pyramids worth of mass per
year. If that rate of loss has been consistent since Pluto was formed, this cor-
responds to between 1 and 10 km of nitrogen ice from its surface that has
potentially been lost.
This is an extremely high loss rate, much higher than expected. It poten-
tially tells us much about Plutos surface. One possible requirement for such
a loss rate is that there is not a lot of nonvolatile material mixed with the
nitrogen and methane we see on Pluto. If there were sand or rock present, it
would gradually come to dominate the surface as the nitrogen escaped.
Eventually, no nitrogen would be present at Plutos surface, and it would
only be found beneath a crust, or lag deposit of other material. Some nitro-
gen might be able to escape by making its way through the lag deposit, but
that would only thicken the lag deposit with time. Because we see such a
high rate of nitrogen loss at the present day, we can infer that there is no lag
deposit present on Pluto.



The flood of TNO discoveries in the late-twentieth and early twenty-first

centuries postdates the discovery of Plutos atmosphere. Given what we
know about Plutos atmosphere, it is natural to apply the same logic to the

TNOs and hypothesize which ones may have atmospheres. As with the first
predictions for Pluto, this work utilizes the spectral properties of TNOs to
figure out their surface compositions. At the temperatures found in the
outer reaches of the solar system, water ice is stable even under conditions
of very low gravity. Compounds like nitrogen (N2), methane (CH4), and
carbon monoxide (CO) can be retained by large objects, but not smaller
Eris has a high albedo and evidence of large amounts of methane on its
surface. This has been interpreted as evidence for a cycle in which methane
frost can condense and sublime, creating a temporary atmosphere as may
be the case with Pluto. However, unlike the case for Pluto, it is thought that
Eriss atmosphere is currently frozen out on its surface. Sedna, by contrast,
has a relatively low albedo. While the data are very sparse and preliminary,
this has been used to argue that Sedna does not have an atmosphere at any
point in its orbit. Since it is never any closer than 79 AU from the Sun, with
a temperature always below 33 K, this is not surprising.
Data for two dwarf planets, Makemake and Haumea, are more ambigu-
ous. Makemake shows evidence of methane ice as well as other compounds
that can be created when the solar wind and other energetic photons inter-
act with methane. As with Pluto and Eris, the predicted temperatures for
Makemake lead scientists to expect it has an atmosphere during part of its
orbit. Interestingly, nitrogen ice is only seen in very small quantities. This
has been interpreted as due to preferential escape of nitrogen from Make-
make in comparison to other dwarf planets, leaving methane behind.
Haumea shows no evidence of methane or nitrogen at all, and the only
compound identified on its surface is water ice. The explanation for this
seems to be related to its companionstwo satellites and five objects in
very similar orbits around the Sun, forming a dynamical family. The origin
of the dynamical family is expected to be via a giant impact into Haumea,
an impact that may have blown off any methane or nitrogen from the sys-
tem, leaving only rock and water ice behind. This theory also fits the sur-
prisingly high density of Haumea, much higher than seen for Pluto or Eris.
On objects with tenuous atmospheres like TNOs, small areas of low and
high albedos can grow to become large areas. This happens because of the
temperature differences that accompany albedo differencesregions with
low albedo absorb more sunlight and become warmer. As a result, volatiles
in the area experience increased sublimation, and areas adjacent to the
higher temperatures also warm up, themselves. Because volatiles tend to
have high albedos, increased sublimation leads to lowering the albedo of
the warm region. This, in turn, increases the temperature further. The end
result can be large low-albedo regions.
However, with the presence of an atmosphere, some volatiles do not
escape into space. Instead, the molecules can travel to colder places where
they condense as frost (as discussed previously for Pluto). The albedos of
the colder areas rise, which, in turn, cools them further in a mirror image of
Small Body Atmospheres  135

the process for warmer areas. As the seasons change, the warmer and colder
regions may move and albedo patterns may change as a result, depending
on the amount of the spin axiss tilt (or its obliquity). Objects whose obliq-
uity is near zero have little seasonal variation and would end up with per-
manent high-albedo polar caps.
Putting together the composition of the ices stable on TNOs and their
temperatures, scientists estimate that methane atmospheres should be
expected within roughly 55 AU of the Sun, while nitrogen atmospheres
could exist out to 120 AU. The details depend upon the gravity and density
of an object, as well as its obliquity. Regardless of those, however, TNO
atmospheres are not expected beyond 120 AU unless something other than
solar heating and ice sublimation is creating the atmosphere (such as out-
gassing due to radioactive heating or impacts adding heat).


Although not an atmosphere in the sense usually considered for planets, the
coma of a comet fits the definition of an atmosphere. As with some of the
TNOs that we have discussed, comets have volatile ices that are stable as sol-
ids when they are further from the Sun, but become gases as they warm up
closer to the Sun. The coma is composed of these gases. In an important
difference with KBOs, however, the materials in cometary comas (or alter-
nately, comae) are not recycled from season to season but rather escape,
with new gas and dust released during each orbit.
Cometary comas are most well-developed for comets nearer the Sun.
While the TNO atmospheres mentioned previously are composed of gases
like methane or nitrogen, comets near 1 AU have temperatures warm
enough to sublime water ice, which is by far the most abundant of the ices.
The coma size is much larger than the nuclear size, in contrast to the atmos-
pheres of the terrestrial planets, which are very thin compared to the plane-
tary radius. In addition to gas, the coma contains dust lofted from the
nucleus by the escaping gas.
As the gas in the coma expands, it begins to cool rapidly, reaching a mini-
mum temperature of only 3040 K. As it expands, however, the lofted dust
is less able to shield it from sunlight, which eventually begins to reheat it. In
addition to heating, energetic ultraviolet photons from the Sun dissociates,
or breaks apart, the original gas molecules in the coma into smaller forms.
These daughter molecules can be detected spectroscopically and give infor-
mation about their original parent molecules as well as the rate of gas out-
flow from the nucleus. For instance, a molecule of water (H2O) can
dissociate into OH and H, or H2 and O. The daughter molecules can often
themselves be broken down as well, in this case the OH could further disso-
ciate into O and H. In addition to dissociation, ionization can occur, leav-
ing OH- or H2-ions in the coma.

Comets making their first visits to the inner solar system (new comets)
are often extremely active. In the transneptunian region, their temperatures
are 30 K or lower. In the Oort cloud, they are lower still, perhaps colder
than 10 K. As they approach the Sun and heat up for the first time in bil-
lions of years, there is plenty of material to be blown off. They experience,
in effect, a reversed version of the condensation sequence described in
Chapter 5the most volatile ices begin subliming farthest from the Sun,
followed by water ice sublimation. Older comets, those that have spent a
longer time in the inner solar system, have largely lost their volatile ices and
have comae dominated by water, relatively speaking.
Because it takes time for heat to penetrate into the interior of an object,
and because new comets have a large reservoir of volatiles near their surfa-
ces, activity can sometimes be seen for new comets even when they are past
perihelion and heading farther from the Sun. Comet Hale-Bopp, for
instance, still had a visible coma and tail even a decade after its perihelion
and after it had retreated past the orbit of Uranus (20 AU from the Sun).
Breakups of comets can also be responsible for new activity. Comets
Shoemaker-Levy 9 (SL9) and Schwassmann-Wachmann 3 (SW3) both
turned into chains of comets as they broke into their constituent pieces and
parts of their interiors saw sunlight for the first time. The composition of

When Comet Schwassmann-Wachmann 3 (SW3) broke up, fragments of its

Figure 10.1
interior were exposed to sunlight for the first time. This caused ice to quickly sublime
and create a coma and tail for each fragment. As a result, each fragment became a tiny
comet of its own. In this infrared image from NASAs Spitzer Space Telescope, dozens
of fragments are seen, each with its own tail. AP Photo/NASA.
Small Body Atmospheres  137

SW3s coma after it broke up was compared to its pre-breakup coma, and
was found to be the same, suggesting its interior composition was similar to
its former surface composition. This is evidence that the surface of SW3,
and by extension, most comets, doesnt change much with time.


Cometary tails can properly be considered as the result of coma evolution.

The gas and dust present in a coma both interact differently with the forces
present in interplanetary space. The dust grains are small enough that they
can be affected by a force called radiation pressure. Radiation pressure is
caused by a transfer of energy from photons striking a surface. It is an
extremely weak force and is easily neglected in everyday life. It is strong
enough, however, to affect the tiny grains of dust and overcome the weak
gravity of a comet. Once lofted into the coma, radiation pressure acts on
the dust grains to push them away from the Sun. Without the effect of radi-
ation pressure, cometary dust would remain in an orbit around the Sun
very much like their parent nucleus. When adding radiation pressure, their
paths change so as to stretch out behind the nucleus, along the cometary
orbit. This is why dust tails have the appearance they do.

Figure 10.2 The gas and dust in a cometary coma act differently as they are swept out into
the tail. The dust follows separate orbits around the Sun under the influence of its gravity
while being pushed by sunlight, while the gas is influenced only by the pressure of sun-
light and moves directly away from the Sun. This results in two tails for most comets, as
seen in this image. Illustration by Jeff Dixon.

Comets also have another kind of tail. The gas in a cometary coma is not
affected by radiation pressure. Ionized gas can be affected by other forces,
however. The solar magnetic field creates a force felt by charged particles,
which can be stronger than gravity. The ionized gas ejected by a comet
moves in response to the solar magnetic field, creating an ion tail or gas
tail, which typically is directed away from the Sun. The ion and dust tails of
a comet may be aligned with one another to a greater or lesser extent, but
they are commonly separate, and differences in color can be seen by a
sharp-eyed person.


The Centaur Chiron is in many ways the largest known comet. It was first
discovered to have a coma in 1989, and its brightness has varied with time,
suggesting outbursts of activity similar to comets. As a result, it was given a
cometary designation (95/P Chiron) to go along with an already-given as-
teroid number (2060). Therefore, it is one of the few objects classified as
both a comet and an asteroid.
After its coma was detected, Chiron was observed frequently through its
perihelion in 1996 and beyond. Confusingly, the observed outbursts have
not been correlated with distance from the Sun (and presumably tempera-
ture) in a consistent way. Water ice has been detected on Chirons surface,
however, even at its hottest surface temperature it will remain solid. As was
discussed in Chapters 8 and 9, the sublimation of near-surface water ice
is thought to be responsible for most cometary activity. CO and CN
have been detected in Chirons coma, and have much lower sublimation
temperaturesas low as 25 K for carbon monoxide. Theoretical studies
suggest that the sublimation of CO may be responsible for Chirons activity.
An additional possibility is that temperature-dependent changes in the type
of water ice in Chiron can lead to outbursts. As Chiron moves toward aphe-
lion in 2021, it will continue to be monitored. In addition to Chiron, there
is one other Centaur known to have a coma, 2001 T4. Little else is known
about that object at this time.


As discussed in Chapter 9, it is thought that many asteroids could have sig-

nificant amounts of water ice in their interiors. In theory, that ice could
sublime if it became warm enough, just as newly exposed ice on a cometary
surface warms and gives rise to a coma.
In 2006, researchers at the University of Hawaii discovered a handful of
objects on asteroidal orbits that have cometary tails and appear to have
comae. Dubbed main-belt comets or activated asteroids because of their
nature, these objects are the best evidence that comets and asteroids grade
Small Body Atmospheres  139

into one another rather than being compositionally unrelated. The known
main-belt comets are members of the Themis dynamical family, found
roughly 3.1 AU from the Sun. While they are too small for detailed compo-
sitional study to be done at present, as members of the Themis family, they
would be expected to be low albedo, C-class asteroids analogous to carbo-
naceous chondrite meteorites (see Chapters 4 and 7). The comae and tails
of these objects appear to be transitory, consistent with an impact excavat-
ing an ice layer from beneath the surface, which then moves in and out of
sunlight with changing seasons, corresponding to times when a coma is
seen or absent. As discussed in Chapter 8, a lag deposit would be expected
to form on main-belt comet surfaces just as on comets. Within a few years,
we will know what fraction of Themis family objects are activated asteroids.
Using that fraction, and estimates of how quickly lag deposits form, scien-
tists will be better able to understand how frequently impacts occur in the
Themis family.


Interestingly, Ceres has also shown evidence of outgassing. Spectroscopic

observations of Ceres in the early 1980s suggested that there might be water
ice on its surface, though at the temperature expected on Ceres, it would
not be stable for long periods. This led to a search for evidence of ice subli-
mation using the International Ultraviolet Explorer (IUE) satellite. The
results were not definitive, but hinted that OH might be present near the
pole of Ceres then experiencing summer, while the other pole showed no
such excess of OH. This was interpreted as consistent with ice heating up
and subliming off of the summer pole, then dissociating into OH and
hydrogen. However, this observation has never been confirmed.
Subsequent findings about Ceres have not favored the presence of water
ice at its surface, as discussed in Chapter 7. However, the discovery that Ce-
res may have a vast internal ocean of ice opens the possibility that water
could potentially be outgassing and Ceres might have a tenuous atmos-
phere. The definitive word on an atmosphere for Ceres will likely need to
await the arrival of the Dawn spacecraft in 2015.


The main-belt comets and Ceres are unusual asteroids because they have
near-surface ice that can sublime to create an atmosphere, like the TNOs
discussed earlier. The vast majority of asteroids have rocky surfaces, which
are stable and do not sublime. Other than the objects mentioned previously,
therefore, no main-belt asteroids have been seen or suspected to possess an
atmosphere. However, it is possible that some of the largest ones like Vesta
could in theory have what is called a ballistic atmosphere, where the density

of molecules is so low that they are much more likely to hit the surface or
escape than they are to encounter another atmospheric molecule. This type
of atmosphere is found on Mercury, the Moon, and several large planetary
satellites. In those cases, the atmosphere is thought to originate from the so-
lar wind knocking atoms like sodium or potassium out of surface rocks (or
knocking a hydrogen or oxygen atom out of an ice molecule on an icy
body). Those atoms can escape quickly or bounce around a few times,
depending upon their energy. Another possible origin for these atmos-
pheres is outgassing of material from an objects deep interior, as could
be occurring with ice on Ceres. The total atmospheric pressures for these
atmospheres are very small, about 1012 times less dense than the sur-
face pressure here on Earth. Such an atmosphere would be difficult to
observe using ground-based equipment, but observations from Vestas orbit
might be able to detect one. Smaller objects do not have gravity strong
enough to maintain even these exceedingly thin atmospheres.


Unlike the planets, the small bodies of the solar system do not have thick
atmospheres. However, there are thought to be thin atmospheres on several
of the large TNOs. The only one of these TNO atmospheres directly
detected is Plutos, which is generated and maintained by sublimation of ice
from its surface. Our expectations of other TNO atmospheres are based on
comparison to Pluto and theoretical calculations. For some objects, the
presence of an atmosphere is indirectly inferred. The sublimation of ice
from comet surfaces creates a temporary atmosphere, which we see as
comae and tails. The gas and dust in these atmospheres escape, requiring
further replenishment. Asteroids do not have atmospheres in general,
although a class of main-belt comets has comae and tails like normal
comets, indicating the presence of near-surface ice. The study of small body
atmospheres gives insight into the conditions at their surfaces.


A more detailed discussion of Plutos atmosphere, touching on other TNO atmos-

pheres, can be found at this Web site, which also has sections focusing on small
body satellites:
Dr. Anne Sprague maintains a set of information about the atmosphere of Mercury
at this Web site. Although it is not an asteroid, Mercurys atmosphere is similar
in character to what one might hypothetically expect on Vesta, if conditions
were right:
The codiscoverer of the first Kuiper belt object, Dr. Dave Jewitt, has written a set of
Web sites about comets and TNOs, including two focusing on the coma and
tails: and http://www.ifa.hawaii.
Small Bodies and Hazards

While most of the small bodies of the solar system are hundreds of millions
of kilometers or more from the Earth, some have orbits that bring them
much, much closer. Occasionally, the Earth is hit with something large
enough to penetrate the atmosphere, sometimes with dire consequences for
life on this planet. In this chapter, we discuss the hazard posed to civiliza-
tion by asteroid and comet impacts, and the initial steps humanity has
taken to try and understand and defend against this threat.


The Earth has been peppered by impacts throughout its history. While ero-
sion by weather and tectonic processes have erased most of the craters it has
accumulated over billions of years, over 170 craters with diameters from
300 km down to a few dozen meters can still be found. Roughly 54 tons of
mass falls on the Earth per year, with objects up to 3 m diameter falling in a
typical year. The largest impactor in historic times occurred in 1908 over
Siberia, an airburst of a 50-m-diameter object. For comparison, the
Sikhote-Alin fall of 1947, the largest meteorite fall recorded, involved an
iron impactor of only 10 m diameterroughly half the size of a railroad
car. Typically, chondritic impactors smaller than 100 m or so do not reach
the ground intact, and objects only a few meters across or smaller com-
pletely burn up in the atmosphere. As objects get larger (or denser), the
atmosphere has less ability to slow them down. Impactors of 100 m or
larger hit the ground in hypervelocity impacts, with speeds faster than the


speed of sound in rock. Impacts at these speeds have more in common with
explosions than a rock dropped from a skyscraper, or even from an air-
plane. The impactor often vaporizes, leaving little of it behind.
The consequences of impacts were first realized in the 1980s. One con-
tributing factor was research into nuclear winter, which was a hypothe-
sized change in climate that could be caused by a large-scale nuclear
exchange between the United States and Soviet Union. The possible climate
change would be driven by large amounts of smoke and soot blocking sun-
light from reaching the surface. It was realized that these same conditions
could be created by an asteroid or comet impact rather than a nuclear
The most influential research in focusing scientific attention on NEO
impacts, however, was the proposal that such an impact was responsible for
the extinction of the dinosaurs. This occurred as the Cretaceous Period
ended and the Tertiary Period began, called the K-T boundary in the rock
record (the traditional abbreviation for the Cretaceous is K). The proposal
that the extinction was caused by an impact was based on an iridium-rich
layer of rock found at the boundary. As discussed in Chapter 4, iridium is a
siderophile element and is concentrated in the Earths core, rarely found in
surface rocks. As a result, the abundance of iridium in this rock layer was
interpreted as extraterrestrial in origin, brought in by an impactor. Soon
after, a crater of the correct age was found on the Yucatan Peninsula of
The recognition that impacts can lead to extinctions helped spur the first
dedicated searches for near-Earth objects. These searches, discussed in more
detail later, along with modeling of impacts, have provided a general sense
of how often collisions occur with various-sized NEOs, and how bad those
impacts are for life on our planet.
While large enough impacts can devastate life on Earth by changing the
climate and starting wide-ranging fires, smaller impacts can also potentially
have effects that cause regional destruction. Small impacts occur much
more frequently than large impacts, although very small objects burn up
in the atmosphere and are not dangerous. Objects a few meters in size
reach the ground as meteorites, but are slowed by the atmosphere to the
point that they are no longer hypervelocity impactors, and do no more
damage than a rock dropped from a great heightpotentially bad for any
unlucky person who might be struck by one (or who owns a car or house
struck by one), but not a general problem for the community.
An important facet of current research is trying to determine more pre-
cisely the size at which impactors become hazardous. This is complicated
by the fact that scientists are still working on the effects of impacts into
water. The most worrisome aspect of ocean impacts is the possibility for a
tsunami, which could affect large areas, some of which are heavily popu-
lated. Computer simulations of tsunamis are generally not designed to
study impact-generated events, which can have very different characteristics
Small Bodies and Hazards  143

than those generated by earthquakes. In addition, most existing simulations

consider those areas where tsunamis have commonly been seen (for
instance, the Pacific or Indian oceans, starting along earthquake faults),
while an impact-generated tsunami could theoretically happen anywhere,
including the Atlantic Ocean or even the Mediterranean Sea.
In September 2007, a crater-forming impact on Earth was recorded for
the first time in history. Near the town of Carancas, Peru, a fireball was fol-
lowed by an explosion that broke windows 1 km from the impact site, leav-
ing a crater 5 m deep and 13 m across. While the original size of the
impactor was relatively small, the high elevation at the impact site (over
3,800 m) meant that the screening effects of the atmosphere were not as
large as they are at sea level.
Remarkably, only a year later, the first Earth impactor to be discovered prior
to impact was found2008 TC3, an object about the size of a compact car.
Roughly a day before impact, 2008 TC3 was discovered, and it was quickly
realized that it would impact near the Egyptian-Sudanese border carrying
roughly a kiloton of TNTs worth of energy, creating an extremely bright
bolide (or fireball) upon entry. Scientists were uncertain if pieces of 2008 TC3
would reach the ground as meteorites, and because of the remote impact loca-
tion, recovery of any such meteorites would be difficult. However, meteorites
were successfully recovered in early 2009, and analyses are under way.

The Tunguska Event
The largest impact in recorded history occurred 7:40 AM on June 30, 1908, in Siberia. For decades,
however, nobody quite knew what had happened. Seismic stations throughout Europe recorded
the blast, and the fires that resulted put enough smoke in the atmosphere to be noticed by North
American observatories for several months.
The remote location of the impact in combination with the upheaval that plagued Russia for
the following decades meant that the blast site could not be visited for some time. When the first
scientific expedition reached the Tunguska site in 1927, they found an area of devastation 50 km in
diameter, though interestingly no crater. Studies of the impact site show the blast had a power of
several megatons, as much as 1,000 times the power of the Hiroshima atomic bomb. It is fortunate
that the impact occurred in remote Siberia, rather than over a city, which would have been a
Before the physics of impacts were well-understood, it was not recognized that an asteroid
impact could have such devastating results, and well into the late twentieth century alternate
explanations were concocted for the Tunguska Event, as it was called, including collision with a
mini black hole or a crash of an extraterrestrial spacecraft. Consensus now has settled on the
more prosaic explanation of an asteroid (or, less likely, comet) impact. Impacts the size of the
Tunguska Event are expected on Earth roughly every thousand years.

Figure 11.1 Tunguska site. AP Photo.


The first near-Earth object (NEO) to be discovered was 433 Eros, in 1901.
At the time, its orbit was seen as a curiosity, though it was recognized that
Eross proximity to the Earth could be used to better calculate the distance
between the Earth and the Sun. Additional NEOs were found over the next
several decades, though the first dedicated searches for them were not
begun until the last quarter of the twentieth century.
The passage of 4581 Asclepius by Earth in early 1989 at roughly twice the
distance of the Moon spurred studies of how to determine the near-Earth
object population. The potential consequences of an impact were dramati-
cally shown in 1994 when Comet Shoemaker-Levy 9 hit Jupiter. This set of
impacts resulted in spots visible in Jupiters atmosphere for months after-
ward in even small telescopes, underscoring the energy involved in the
impact of even small objects.
In 1998, the United States Congress mandated NASA to find 90 percent
of all NEOs with sizes 1 km diameter or larger, capable of causing global ca-
tastrophe, within 10 years. NASA has funded the Spaceguard Survey as a
result, composed of several independent surveys, each with specific
strengths. As of the closing months of the Spaceguard Survey, astronomers
estimate they will come close to surveying the NEO population to the
requested level. The first searches used technology similar to that used to
discover Pluto in the 1930sphotographs of the same patch of sky taken at
different times and visually compared to one another using a blink compar-
ator, which is a machine that rapidly switches back and forth between two
photographs as seen through an eyepiece. Objects like stars remain fixed
between the two frames, while moving objects jump back and forth between
Small Bodies and Hazards  145

their positions on each photo. This technique is still used with digital im-
agery on computer screens, where it is called blinking. Newer search pro-
grams developed software that could automatically identify every object on
an image and compare it to images of the same area at different times. After
removing all of the objects whose positions are identical on the different
images (and thus are fixed stars), the remaining objects are possible NEOs
(or main-belt asteroids or comets). Current techniques require at least three
images to be observed to allow the software to determine if any potential
NEOs are moving in a manner consistent with a real orbit.
While different search programs have differing strengths, all must deal
with certain observational limits and difficulties. First is that NEOs can go
through phases, just like the Moon. As a result, they are brightest when their
phase is full and the Earth is directly between them and the Sun. This
alignment is called opposition. Further from opposition, objects become
fainter. Objects at opposition are always highest in the sky at midnight, op-
posite the Sun. Many search programs therefore concentrate on that part of
the sky, knowing that any NEOs they find will be nearly as bright as they
can be.
Light from the Moon interferes with observing nearby. As with NEOs,
full moon occurs at the opposition point in the sky, and observing fainter
objects nearby becomes difficult to impossible. As a result, many search
programs choose not to observe at all near the time of full moon. The Milky
Way is also a difficult area to observe, due to the very high concentration of
stars and the likelihood of confusing a faint object with a background star.
Distant objects like Eris and Pluto move so slowly that their opposition
dates change very slowly from one year to the next. Objects closer to the
Earths orbit have opposition dates that change more rapidly. The time
between oppositions is called the synodic period. This can be thought of as
though the Earth and other solar system bodies are cars on a racetrack. The
synodic period is how long it takes Earth to lap other objects, and the more
similar an objects speed is to the Earths, the longer it takes Earth to lap
them. The synodic period of many NEOs can be very long, perhaps decades.
Because the objects with orbits most similar to the Earth are also the ones
most likely to impact the Earth, and because their synodic periods can be so
long and they spend so much of their time far from opposition, some
search programs have begun to include search areas far from the opposition
Figure 11.2 shows a simplified version of the situation for Earth and the
NEO 10563 Izhdubar, with the relative positions of Earth, the Sun, and Izh-
dubar projected onto the ecliptic plane on March 21 of every year. Objects
to the right of the vertical line are visible during the night, and objects to
the left are visible during the day, though close-to-the-line objects can be
visible at night close to dawn or dusk. In 1993, Izhdubar was discovered,
and in 20092010 it will be opposite the Sun from Earth. Izhdubars year is
roughly four days longer than the Earths and it will slowly fall behind the

Figure 11.2 This diagram is a snapshot of the positions of the Earth, Sun, and the asteroid
10563 Izhdubar every fifth March 21 from 1995 to 2090. Because it is the same day each
year, the Earth is always in the same spot relative to the Sun. Izhdubar, however, is found
in very different parts of the sky on March 21 as the years pass. The vertical black line
separates the times when Izhdubar is visible during the daytime (to the left of the line)
and at night (to the right of the line). As can be seen, for the vast majority of the 100-year
period shown, Izhdubar is not visible at night. The same is true of many other NEOs,
which has led search programs to try and counter this effect by proposing space-borne
telescopes which could search much closer to the Sun. Illustration by Jeff Dixon.

Litter in Near-Earth Space
The vast majority of material outside Earths orbit is transported from the main asteroid belt.
However, humanity has sent a number of space probes to explore the solar system. While the main
probes are tracked, booster rockets, instrument covers, and the like are often jettisoned and
untracked (or untrackable). As NEO surveys become sensitive to smaller and fainter objects, some
of these objects will be detected and catalogued, though definitive identification could be difficult.
A dramatic example of this was seen in 2007 as the European Rosetta spacecraft was preparing for
a gravity assist using the Earth. Ground-based surveys identified it and it received a provisional desig-
nation of 2007 VN84 before suspicious observers noted the similarity between its orbit and Rosettas.
In 2002, an object with an orbit very similar to Earths was discovered and named J002E3. Astrono-
mers (including your humble author) obtained its reflectance spectrum, and found that it matched
white titanium-based paint, clinching a human-made origin. This was supported by orbital calcula-
tions that showed J002E3 had last made a close approach to Earth in 1971. The current best guess for
the identification of J002E3 is an upper stage for Apollo 12. J002E3 is once again wandering in inter-
planetary space, and is expected back in the neighborhood in roughly 2032.
Small Bodies and Hazards  147

Earth as they both make their way around the Sun. For most of Izhdubars
orbit, it is only visible during the day, spending decades at a time close to
the Sun (for instance, between 2030 and 2060). However, it also spends sev-
eral decades visible shortly before sunrise (for instance, the period from
2075 to 2095 and beyond). This type of behavior is true in general for
NEOs, not just Izhdubar. These areas near sunrise and sunset have infor-
mally been dubbed the sweet spots by asteroid searchers and are prime
targets areas for the newest NEO searches. The sweet spots are particularly
important when observing those objects whose orbits are largely or entirely
within the orbit of the Earth. Such objects spend very little time visible in
the nighttime sky, making them particularly difficult to discover.



Our current understanding of the NEO population and its threat is shown
in Figure 11.3. There are estimated to be roughly one thousand objects of
the size capable of causing global devastation with an impact. None of the
ones we know of are on orbits that will impact the Earth. The size-fre-
quency distribution (a concept first discussed in Chapter 6) of NEOs leads
us to expect roughly 50,000 objects larger than 140 m, which is a size that
could cause regional devastation, with an impact energy the equivalent of a

Figure 11.3 The size frequency distribution of NEOs, in combination with the average
speeds with which they impact the Earth, can be used to estimate the frequency of
impacts of different sizes. Larger impacts occur rarely but have devastating consequen-
ces. Illustration by Jeff Dixon.

typical nuclear weapon (510 megatons of TNT). As mentioned before,

smaller objects do not usually make it through the Earths atmosphere
unless they are the relatively rare objects made of iron. The calculations used
in the figure are based on objects with the density of rock rather than ice or
metal. The denser the impactor, the bigger the energy of the impact and the
more destructive it is. An impact speed typical of asteroids is also used.
The role of comets in the impactor population is not completely under-
stood. They are believed to be a relatively infrequent threat compared to the
asteroids. In addition, searches for hazardous comets are beyond the capa-
bilities of current search programs. A cometary impact is likely to be at
extremely high speed, potentially up to four times faster than a typical as-
teroidal impact. As a result, the warning time might be exceedingly short.
While cometary impacts are currently not considered to be preventable, it is
likely that within the next few decades the largest remaining uncharacter-
ized risk will be from their potential.
Given the large number of potentially hazardous asteroids and the fact
that orbits are not immediately known to high precision, it is not surprising
that some objects have a non-zero impact probability shortly after their dis-
covery. Because this is an unfamiliar threat to many, coverage of these situa-
tions in the media has occasionally been confused.
The increasing number of NEOs that have been discovered have led to a
corresponding increase in the number of possible impactors that have been
found. Additional data for these objects, whether newly acquired or
archival, has always improved knowledge of their orbits to allow an impact
to be ruled out. However, before this additional data has become available,
the threat of impact is not zero. The most dramatic case of this was in late
2004, when the object 99942 Apophis, then known by its provisional name
of 2004 MN4, was thought to have a 1 in 38 chance of hitting the Earth in
2029 before new data showed there would be no impact. There have been
some cases where the media has publicized a possible impact while the cal-
culations were very uncertain and additional data were being gathered. Sci-
entists have weighed the public right to know about any potential collision
with the desire to present a complete, considered picture to the public and
avoid any panic. And unsurprisingly, differing personalities lean toward dif-
ferent ends of the argument.
Part of the effort to better inform the public of the risk of NEO impacts
has been the creation of two measures of the threat from any specific aster-
oid. The Torino Scale (Figure 11.4) categorizes objects based on the likeli-
hood of impact and the consequences if an impact occurred, ranging from
0 for objects that will not hit the Earth or will burn up in the atmosphere,
all the way to 10 for objects that are civilization-threatening. Apophis
reached 4 on this scale before new data allowed it to be recategorized as a 0.
The object 2008 TC3 was sufficiently small that even though it impacted
the Earth, it never reached more than 0 on the Torino Scale. The Palermo
Scale has been proposed as an alternative to the Torino Scale. It is more
Small Bodies and Hazards  149

Figure 11.4. The Torino Scale is an attempt to inform the public about the hazards posed
by specific objects based on their odds of impact and the energy that would be released
if they did impact (which is related to their size). As orbital information improves for
an object, it can move between categories. The bottom panel provides an explanation
of each category. Every known object is either a 0 or 1 on the Torino Scale. The NEO
99942 Apophis briefly reached a 4 on the Torino Scale before additional data moved it
back to 0. This was the highest any object has yet reached. NASA.

technical, and includes the time until possible impact as part of the calcula-
tion. While the Torino Scale only includes integers, the Palermo Scale allows
for fractions. The Palermo Scale also includes a comparison to the back-
ground probability of an impact, which is the chance Earth will be hit by
something currently undiscovered. Again, Apophis has had the largest
Palermo Scale value in history, topping out at 1.1 before falling to 2.5.
Because the Palermo Scale is a logarithmic scale, this is read that Apophis at
its peak was 101.1 (or about 13) times more likely to impact than a currently
undiscovered object, but now is only 10 2.5 or roughly 1/300 as likely.
Hazards in Space
Impacts do not only occur on Earth, of course. Much of our knowledge about impacts comes from
studying their effects on the Moon and other planets. The only large impact ever witnessed was
that of Comet Shoemaker-Levy 9 into Jupiter during the mid-1990s, though for some time in early
2008 while its orbit was uncertain it was thought that the NEO 2007 WD5 had a chance of impact-
ing Mars.
Until relatively recently, humanity has not had to worry much about non-Earth hazards. With
the International Space Station (ISS) constantly staffed with astronauts and cosmonauts and the
possibility of lunar bases in coming decades, however, scientists have begun to calculate the likeli-
hood of impacts affecting them.
Unlike people living on Earth, space stations and lunar bases do not have the benefit of an
atmosphere to screen out small, fast-moving micrometeorites. Furthermore, scientists do not
have a precise idea of how many of the smallest micrometeorites exist in space. And the conse-
quences of even tiny impacts could be severe if they are unluckily placed or puncture pressurized
areas. There have been some events, for example, the old Soviet/Russian Mir space station suf-
fered four small impacts in 15 years, though none were serious. A study conducted in 2007 esti-
mated that the ISS had as much as a 5 to 9 percent chance of being destroyed by a meteoroid
Lunar bases under consideration also run similar risks. However, they have some potential
advantages that space stations dont have. The most obvious advantage is that they can be built
partially or wholly underground, which would protect them from all but relatively uncommon
large impacts. Conversely, the cheapest and fastest construction for a lunar base would be using
inflatable habitats, which would be much more susceptible to micrometeorite damage than
the ISS.


Along with the efforts to catalogue the near-Earth and potentially hazard-
ous object populations, there has been research into the best way to handle
an object on a collision course. These mitigation techniques vary from rela-
tively gentle to more violent. Not surprisingly, the gentler techniques
require longer times to be effective. At the most basic level, mitigation ideas
can be separated into impulsive techniques, which attempt to deflect the
incoming object with a single blow, and continuous techniques, which
nudge the target without running the risk of disrupting the target. The most
efficient way of changing the would-be impactors orbit is by slowing or
speeding it upthe Earth moves a distance equal to its radius in under four
minutes, so a change in arrival time by that amount should allow an impact
to be avoided.
The most dramatic of the impulsive techniques is the use of a nuclear
weapon (or several nuclear weapons) to deflect the target. Contrary to the
conception of this method often shown in movies, the weapons would not
be used to blow up the incoming body, which could potentially lead to
Small Bodies and Hazards  151

multiple collisions by the resulting fragments and leave us worse off than
originally. Instead, the nuclear weapons would be detonated at some dis-
tance from the body, and the absorption of neutrons from the explosion
would impart momentum to the NEO and move it.
Many people are exceedingly uncomfortable with the use of nuclear
weapons for mitigation, even for the cause of averting catastrophe. There is
some fear that acceptance of their use, even in a good cause, would lead to
the militarization of outer space. Others argue that a large number of
launches may be required to ensure that an incoming object is deflected,
and the consequences of a launch failure for a nuclear weapon could be dev-
astating. Nevertheless, the general consensus among researchers is that if a
large object is found a short time before impact in the near future, deflec-
tion via nuclear weapons may be our only realistic response.
The other major impulsive technique is use of a so-called kinetic
impactor. A kinetic impactor is simply a mass sent to collide with a target
with a speed and direction to impart the required momentum to move
the target away from a collision course. It is effectively the same tech-
nique used in billiards. Kinetic impactors are attractive because they
could provide a straightforward means of deflection that does not use
nuclear weapons. However, the details could be difficult to work out since
unlike billiard balls, NEOs will create ejecta when they are hit, which also
will carry momentum and affect the final trajectory of the target. Further
research has been proposed to better understand how kinetic impactors
can be used, and practice missions have been suggested as a means to
do so. It is likely that this technique will be the first one tried out as a
Continuous techniques cause small changes in target velocity at any
given time. They are considered to be better suited to smaller objects, and
particularly objects that do not require large orbital changes to miss the
Earth, particularly objects for which potential impacts are far in the future.
Continuous techniques have the benefit of allowing constant monitoring of
the impactors orbits, without any of the uncertainties associated with im-
pulsive techniques. Conversely, continuous techniques can potentially
require spacecraft to operate for very long times, and a failure partway
through a mission may end up simply moving an impact to a different loca-
tion on Earth. Finally, continuous techniques require a rendezvous with the
NEO to be deflected, while impulse techniques can potentially be used on
less-expensive flyby missions (although targeting on flyby missions will
potentially be challenging).
Perhaps the most obvious of the continuous techniques is literally push-
ing the impactor out of Earths way. In the simplest case, this would involve
a spacecraft attaching to a NEO and firing a rocket. This type of mission,
sometimes called an asteroid tug, would need to account for the rotation
of the target, to not simply increase its spin rate rather than moving it. Pro-
ponents of the asteroid tug have suggested that a few months of continuous

operation roughly a decade before any potential impact date would be suffi-
cient to deflect a typical threat.
The mass of an impactor can be estimated given a density and size. For a
100-m non-porous rocky body, this works out to roughly 1.5 billion kg. For
comparison, the Space Shuttle orbiters have a mass of 100,000 kg. While
this is only a small fraction of the NEOs mass, if the Space Shuttle were
close enough to a 100-m NEO, it would exert a gravitational acceleration
on it. With enough time, a threatening NEO could in theory be deflected
simply by putting a massive spacecraft nearby and dragging the NEO
using the spacecrafts gravity. This concept, called the gravity tractor, has
been proposed by some as an ideal way of handling threats from small
NEOs. It has the benefit of not requiring any attachment to the target and
being insensitive to the rotation rate or spin direction, in contrast to the
asteroid tug. However, it requires the tractor to be in operation at a very
small distance from the target, which might be difficult to manage. The
gravity tractor is another concept that may be tested in the relatively near
Additional mitigation ideas have been proposed, but some initial stud-
ies show them to have greater issues than the ones already discussed. The
most notable of these ideas are the use of a mass driver and changing
the objects albedo. A mass driver removes material from the target sur-
face and ejects it at high speed. This, due to conservation of momentum,
would slowly propel the object in the desired direction. However, rotation
of the target must be dealt with in the same way as for the asteroid tug.
Also under study has been the possibility of taking advantage of the
Yarkovsky Effect by changing the albedo of a threatening object, perhaps
by coating it with coal dust. If done cleverly, the change in albedo could
change the targets thermal properties so that the Yarkovsky Effect would
move it out of a collision course with Earth. As with the other continu-
ous techniques, this would require a long lead time to be effective,
although in contrast with the other techniques, it would not require the
continuous operation of a spacecraft.
Although not mitigation per se, there is also the option to do nothing
and allow an impact to occur. This will be the response for some predicted
impacts, for instance for those objects too small to have an appreciable
effect. It might also be determined that an impact will occur in the ocean
and the most cost-effective response will be evacuation of coastal areas
rather than attempting to stop the impact.
Beyond the technical questions, there are also many policy and legal
issues that potentially complicate any mitigation efforts. It is not clear
whether an international body like the United Nations would claim juris-
diction over potential mitigation efforts, or if individual spacefaring coun-
tries would act independently or in consortia. From a financial liability
standpoint, it is not clear whether the responsibility for an unsuccessful
mitigation effort would be assigned to any party, and whether NASA or the
Small Bodies and Hazards  153

United States, for instance, could be sued. Along similar lines, thresholds
above which action would be taken are also not establishedif a mid-
Pacific impact would devastate some sparsely populated islands, it is not
clear whether the proper response would be to attempt mitigation or simply
evacuate the area.
There is also no consensus for a threshold probability at which action will
be taken. From a mitigation perspective, the earlier action can be taken the
less force is required to succeed. On the other hand, the uncertainties in
impacts are often large enough that the estimated probabilities of impact
are only a few percent until several months ahead of time. If mitigation mis-
sions are launched each time the probability of impact reaches 2 percent,
for arguments sake, 49 of 50 will have been unnecessary. Given the large
cost of mitigation missions, and the risks present in the missions them-
selves, this presents a quandary.


The Earth has been impacted by asteroids and comets uncounted times in
its history, up to the present time. These impacts have been responsible for
large, wide-ranging extinctions. Ongoing surveys of near-Earth objects have
discovered nearly a thousand bodies large enough to have a drastic effect on
human civilization if they impacted, though no objects on collision courses.
New surveys will focus on smaller objects capable of regional devastation.
Scientists and engineers have identified a number of techniques for possible
prevention of impacts, though some are controversial. Impulsive techniques
are the ones that are closest to readiness, and thought to be the most appro-
priate in cases of little warning. Continuous techniques provide more con-
trol over the outcome, though require technology that may not be ready for
some time. Beyond technological issues, political and ethical questions sur-
rounding NEO impact and mitigation remain largely unasked, let alone


The issues of near-Earth objects and the impact hazard they pose are the subject
of several compilations on the Internet. The Planetary Society has a set of
popular-level links at this URL:
The Minor Planet Center is the central clearinghouse for collecting NEO discov-
eries and observations and calculating orbits. This Web site has information
about these observations, including technical information about astronomers:

The Jet Propulsion Laboratory (JPL) is in charge of NASAs effort to track and
characterize NEOs and has a wide array of information at its Web site: http://
Also at the JPL Web site is the text of a recent report to the United States Congress
compiled by experts:
This site has a different look at the impact hazard by scientist and artist Bill Hart-
mann; it details his attempt to make a painting of the Tunguska Event as scien-
tifically accurate as possible:
Spacecraft Missions

Only in the last half-century has humankind been able to visit other mem-
bers of the solar system. And only in the last quarter-century have small
bodies been explored close-up by spacecraft, with the first visits to dwarf
planets planned for the coming decade. Spacecraft missions have revolu-
tionized our knowledge of asteroids and comets, and planned missions will
further our knowledge even more. In this chapter, we will look at the vari-
ous types of spacecraft missions that exist and the instrumentation they
have carried (or will carry), consider the unique benefits and limitations of
spacecraft missions, and discuss what has been done and what we can
expect in the future.


When considering the ease (or difficulty!) of reaching a particular target

object, spacecraft navigational engineers must take the difference or similar-
ity to the Earths orbit into account. Because the Earth and the target body
are both orbiting the Sun at different speeds, spacecraft must carry enough
fuel to change from the Earths orbit to the targets. The change in speed
required for this orbital change is called delta-v, which is a standard meas-
ure of the relative difficulty of reaching an object. For the Moon, delta-v is
roughly 6 km/s. Some near-Earth objects (NEOs) have delta-v as low as
4 km/s, meaning they require less fuel for a mission than do missions to the
surface of the Moon. Orbital speeds are related to the inverse square root of
the distance to the central body, so an object at 4 AU would have an orbital


speed roughly half that of Earth, while one at 100 AU would be traveling
roughly one-tenth as quickly as the Earth. This means that it is relatively
easier in terms of fuel consumption to reach outer solar system bodies than
you might thinkin terms of delta-v, it is only about two to three times
more costly to reach the distance of Pluto and even Eris than it is to reach
Mars. In practice, however, delta-v requirements to visit small solar system
bodies are larger than the best-case scenarios because unlike the major plan-
ets, their orbits are seldom near the same plane as the Earth. Changing orbit
planes is very costly in terms of extra fuel required.
The simplest calculation of delta-v, still somewhat more involved than
can be presented here, assumes a particular type of orbit called a Hohmann
transfer orbit. Such an orbit is an ellipse tangent to both the Earths and tar-
gets orbit. For a body like Mars, the transfer orbits periapse is at the
Earths orbit and its apoapse is at Marss orbit further from the Sun. For tar-
gets closer to the Sun than the Earth is, the transfer orbits apoapse is at the
Earth and its periapse is at the target.
The path followed by a spacecraft (also called its trajectory) is often
more complicated than a Hohmann transfer orbit. It was recognized in the
mid 1970s that close passes near planets can be used to deflect spacecraft
trajectories and conserve fuel. Many missions now take advantage of these
gravity assists, often using the Earth. Figure 12.1 shows a schematic of
Dawns trajectory as an example.

Figure 12.1Dawn will visit both Vesta and Ceres, helped along by a close pass to Mars
that will use that planets gravity to alter its orbit. This diagram shows the critical dates
in the mission timeline, along with a schematic view of the positions and orbits of
Earth, Mars, Vesta, Ceres, and Dawns path. NASA.
Spacecraft Missions  157


Missions to small bodies (or indeed to any object) provide the oppor-
tunity to perform studies that are impossible via remote sensing.
Instead of calculating a bodys dimensions by measuring the brightness
changes of a point of light (as described in Chapter 6), the size and
shape of an asteroid or comet can be directly seen. Instead of getting a
reflectance spectrum that measures the average composition, differences
across an object can be measured. Furthermore, some data can be taken
that is impossible to obtain from the Earththe regolith of an asteroid
or the coma and tail of a comet can be directly sampled, or the elemen-
tal ratios of a body can be measured. In the most complicated mission
scenarios, material can be brought back to be measured in laboratories
on Earth.
In addition to these purely scientific reasons for carrying out missions,
there are other motivations that sometimes come into play. Missions can
serve roles as engineering or technology demonstrations. For instance, the
Deep Space 1 mission to Comet Borrelly was largely intended as a test of a
new type of propulsion, which has subsequently been used on other mis-
sions. The exploration of outer space is seen by many as lending prestige to
a nation and as a means of stimulating interest in science and technology
among its population. Missions have also been the source of international
cooperation through the decades, as one nation will often participate in
building instruments for another nation.
To date, there have been four leading players in space exploration
beyond Earth orbit: the United States of America (through its space
agency NASA), Japan (its space agency is called JAXA), the Soviet Union
(before its dissolution in 1991), and the European Space Agency or ESA,
a consortium of 17 European nations (though most of its funding comes
from the United Kingdom, France, Germany, and Italy). Additional
nations are beginning to have interest and capabilities in planetary ex-
ploration, with Russia expected to launch a mission to Phobos in the
coming years and India and China both launching lunar missions and
planning further exploration. In addition, ESA member states Germany
and France have shown some interest in independently undertaking mis-
sions to visit small bodies.


There are several different types of spacecraft missions, varying in complex-

ity and difficulty, as well as in the amount of data return and the instrumen-
tation typically carried. These mission types can generally be described as
flybys, rendezvous, landers, and sample returns. The set of instruments
carried on a mission is typically called the payload.
Table 12.1. Past, Present, and Future Missions to Asteroids, Comets, and Dwarf


Comet Giacobinni- ICE 1985 flyby
Comet Halley Vega 1, 2; Giotto; 1986 flyby
951 Gaspra Galileo 1991 flyby
Comet Grigg- Giotto 1992 flyby
243 Ida Galileo 1993 flyby
253 Mathilde NEAR Shoemaker 1997 flyby
9969 Braille Deep Space 1 1999 flyby
433 Eros NEAR Shoemaker 2000-2001 orbiter, lander
Comet Borrelly Deep Space 1 2001 flyby
Comet Wild 2 Stardust 2004 flyby, sample return
Comet Tempel 1 Deep Impact, 2005, 2011 flyby
Stardust NExT
25143 Itokawa Hayabusa 2005-2007 orbiter, sample
2867 Steins Rosetta 2008 flyby
21 Lutetia Rosetta 2010 flyby
Comet Hartley Deep Impact/ 2010 flyby
4 Vesta Dawn 2011 orbiter
Comet Churyumov- Rosetta 2014-2015 orbiter, lander
134340 Pluto New Horizons 2015 flyby
1 Ceres Dawn 2015 orbiter


The easiest mission type to undertake in many ways is the flyby. A flyby
mission has an orbit that allows a close pass to the target, during which data
are collected. The delta-v requirements can be quite modest, but the en-
counter, or duration of the data collection period, can be short. Flybys were
the first planetary missions that were performed by spacefaring nations,
and have been the means by which initial reconnaissance of objects have
been undertaken. Historical examples are the Voyager 1 and 2 spacecraft of
the 1970s and 1980s, which conducted a series of flybys of Jupiter and Sat-
urn, with Voyager 2 adding flybys of Uranus and Neptune.
Spacecraft Missions  159

Similarly, the first views of the small bodies have been via flyby. The Soviet
Vega 1 mission obtained the first images of a cometary nucleus in 1986 as it
encountered Comet Halley, as did several other international spacecraft, and
the first spacecraft to visit an asteroid, Galileo, did so en route to Jupiter.
Because they spend only a limited time near their target body, flyby
spacecraft typically have instruments suited for quick, remote observations.
Instruments commonly on board flyby spacecraft (indeed on most space-
craft in general) include cameras for imaging and spectrometers to measure
reflectance properties. Some will also carry a LIDAR (light detection and
ranging), which measures the time between emitting and detecting a pulse
from a laser in order to measure distances very precisely.
During Comet Halleys most recent visit to the inner solar system in 1985
1986, several spacecraft (collectively referred to as the Halley Armada)
performed flybys through its tail and coma. These allowed mass spectrome-
ters to be used. Mass spectrometers ionize samples and then measure how
those ions act in electromagnetic fields to deduce their atomic weight. As a
result, the elemental abundances can be measured with some accuracy.
Finally, a close pass by an object can allow its mass to be measured. This
is done by observing the spacecraft orbit before and after the closest
approach and measuring the deflection caused by the object. However, the
precision of the measured mass is a sensitive function of how small the close
approach distance is, as well as how fast the flyby speed is. Slow speeds and
close passes provide much better data than fast flybys and distant passes.
Asteroids are sufficiently numerous that there are often opportunities for
spacecraft flybys while they are traveling to a different destination. As men-
tioned before, the first asteroid encounters were flybys, as the Galileo space-
craft passed the asteroids Gaspra and Ida en route to Jupiter. NEAR
Shoemaker encountered the asteroid Mathilde on the way to a rendezvous
with Eros. Similarly, the Deep Space 1 mission had a flyby of the asteroid
Braille on the way to Comet Borrelly. The CONTOUR mission was designed
to fly by at least two cometary nuclei (with the possibility of a third), but
was lost before its first encounter. More recently, the asteroid Steins was
visited by the European Rosetta spacecraft in mid-2008, with a second flyby
planned before its final destination is reached.
An interesting twist on a traditional flyby mission was taken by the Deep
Impact spacecraft. Shortly before its encounter with Comet Tempel 1, it
released a 370-kg probe with a camera designed to crash into the comet,
while the main spacecraft flew by. The scientific goals were to see if come-
tary activity could be induced and allow crater formation to be observed on
a large scale. At the same time, the comet was observed by a large number
of astronomers on Earth as well as orbiting observatories. The mission was
very successful, although ironically neither of the goals were meta large
amount of dust was created in the impact, which blocked a view of the cra-
ter formation, and an outburst was not started by the impact. However, the
data collected and observations of the dust provided a much greater under-
standing of cometary surfaces (see Chapter 8).

The Deep Impact mission released an impactor and flew by the target comet
Figure 12.2
watching the impact. In this cartoon timeline, the spacecraft is moving from right to
left. NASA.

The Halley Armada
The return of Comet Halley to the inner solar system in 1985 provided an unprecedented opportu-
nity for spacefaring nations. Plans for missions to encounter Halley were in the works by NASA,
the European Space Agency (ESA), the Soviet Union, and Japan by the late 1970s. At that point,
there were hopes for a joint NASA/ESA mission to fly by Halley en route to a rendezvous with
another comet a few years later. Budget pressures caused NASA to drop out of the joint project,
leaving ESA to run a modified version of the mission alone. ESA named the mission Giotto, after a
fourteenth-century artist who included a rendition of Comet Halley in his works. In an effort to
find some way to study Halley, NASA rededicated the International Sun/Earth Explorer 3 satellite
from studying the Earths magnetic field, sending it via a complicated series of maneuvers and
gravity assists to the vicinity of Comet Halley and renaming the spacecraft the International Come-
tary Explorer (ICE)
In addition to Giotto and ICE, there were four other spacecraft that had flyby encounters with
Comet Halley, a group collectively called the Halley Armada: Vega 1 and Vega 2 (both Soviet/
French missions), and Suisei and Sakigake (both Japanese). The encounters all occurred during
March 1986, all but ICE during the period of March 614. The Halley Armada approach allowed
an international division of labor, which increased the scientific return of each mission relative to
its cost.
Ironically, while the Halley Armada was designed to be humanitys first encounter with a
comet, ICE encountered Comet Giocobini-Zinner in September 1985 while en route to the vicin-
ity of Comet Halley. It thus became the first spacecraft to encounter two small bodies. Other
members of the Halley Armada also had additional targets: Vega 1 and 2 both encountered
Venus prior to Halley and dropped probes on that planet, while Giotto encountered Comet
Grigg-Skjellerup in 1990.
Spacecraft Missions  161


With additional fuel and the proper timing, a spacecraft can rendezvous
with its target rather than merely fly by. Smaller objects have weak enough
gravity that a spacecraft will fly in formation with it around the Sun rather
than orbit it. Larger objects such as Vesta have sufficiently strong gravity
that a spacecraft will orbit it. In either case, however, a rendezvous mission
spends a significantly longer time at its target than a flyby does.
Because of this, the amount and types of data that can be returned during
a rendezvous mission are much larger than for a flyby. Images and spectra
can be obtained over the entire surface of the target body instead of just the
side visible during the flyby. Use of a LIDAR combined with knowledge of
the spacecrafts position can provide a very accurate shape over an objects
entire surface, allowing the calculation of a precise volume. Because rendez-
vous missions allow the determination of the targets mass much better
than during flybys, this combination results in good values for the density
for the target.
In addition, other instruments have been carried on small body rendez-
vous missions. Gamma-ray spectrometers (GRS) measure gamma rays (high-
energy photons) emitted by elements when they have themselves been hit
by high-energy cosmic rays. The various elements emit gamma rays with

Figure 12.3 Neutron and gamma ray spectrometers are used by spacecraft to determine
compositions. They take advantage of differences in the way elements react to high-
energy photons to measure the relative concentration of those elements. They have
been common instruments on planetary missions, including NEAR Shoemaker and

different energies, allowing them to be distinguished from one another. The

resulting data give the relative abundances of many important elements
including carbon, oxygen, silicon, iron, and magnesium in the top meter or
so of the targets crust. The rate of gamma ray emissions is relatively slow,
however, and therefore a relatively long time is required to collect the neces-
sary data. Flybys occur much too quickly, but NEAR Shoemaker and Haya-
busa both carried GRS instruments, which showed that Eros and Itokawa
were similar to ordinary chondrite meteorites in composition (see Chapter 7).
X-ray spectrometers (XRS) work similarly, though the photons have different
energies and originate from the Sun rather than cosmic rays. Often a neutron
spectrometer (NS) is also included; these instruments measure the flux of neu-
trons with differing speeds from a surface, which allows the presence and
abundance of hydrogen to be measured. The Mars Odyssey spacecraft used a
NS to measure subsurface water ice on Mars. The Dawn spacecraft is carrying
an instrument called the Gamma Ray and Neutron Detector (GRAND) to
Vesta and Ceres. Rosetta will visit Comet Churyumov-Gerasimenko in 2014
and carries 11 instruments from Europe and the United States to provide
comprehensive information about their target.


An additional step in complexity can be taken by landing on the target.

Landing on objects like Mars or the Moon requires additional delta-v, but
again the weak gravity of the small bodies means that it is not difficult to
change the spacecrafts speed to be able to land. Ironically, however, it is the
weak gravity that causes other difficulties. On Mars or the Moon, the grav-
ity is strong enough that engineers can easily calculate the speed at which
the spacecraft will approach the surface when in free fall, and the spacecraft
can be programmed to fire retrorockets when it reaches a particular speed
in order to give a soft landing. Any small uncertainties in the falling speed
or imprecision in the rocket firing will be a tiny percentage of the total
speed and thus wont have a big effect.
On small bodies, on the other hand, the gravity is small. Escape speed on
Eros is 10 m/s, roughly the speed of an Olympic sprinter. On Itokawa, it is
20 cm/s, much slower even than a comfortable stroll. Landing speeds will
be similar to these escape speeds. At these speeds, inaccuracies of only a me-
ter or so in position, or a fraction of a second in firing a rocket, can spell
the difference between a soft landing and a crash or a bounce. Indeed, the
Hayabusa spacecraft suffered several slow-motion crashes into Itokawa dur-
ing its mission. An additional complication is due to the irregular shape of
small bodies, which results in variations of even this weak gravity depend-
ing on location. Finally, the weak gravity means that thought must be given
to anchoring the spacecraft on the asteroidal or cometary surface, otherwise
the lander might accidentally throw itself off the surface entirely!
Spacecraft Missions  163

Once on the surface, however, additional possibilities are available to

landed spacecraft relative to rendezvous missions. The time required for
collecting useful data with a GRS or XRS decreases with decreasing distance
from the object, so landing on the surface is optimal. Cameras can provide
extremely detailed images of the surface, though they must be designed to
operate in focus at very short distances. Additional instruments can be used
on the soil, either by placing the instrument directly on the surface or by
scooping regolith and bringing it into an experiment chamber inside the
Although these techniques have not been used on a small body mission,
they have been used on landers on Venus, the Moon, and Mars many times.
They are also planned for the Philae lander, which is part of the Rosetta
mission. Philae will harpoon itself to the surface of Comet Churyumov-
Gerasimenko (often called C-G) in 2014, and will also drill itself into the
surface. The harpoons will have instruments of their own, measuring the
density and strength properties of the surface. Philaes instruments will pro-
vide isotopic and mineralogical data about C-Gs surface in much greater
detail than is possible from a rendezvous alone. It will also drill 20 cm below
the surface and study that subsurface material. As with many planetary
landers, the Rosetta mission has a rendezvous craft in addition to a lander.
This strategy allows a smaller lander, and the opportunity to use the rendez-
vous craft to relay communications to Earth, which allows the lander to
only need a small antenna.
The first asteroid lander was NEAR Shoemaker, which was designed as a
rendezvous mission. After its mission objectives were achieved it was
brought in for a landing on a pond on Eros as an experiment, which bril-
liantly succeeded. NEAR was not designed as a lander, and its instruments

NEAR Shoemaker visited the asteroids Mathilde and Eros, carrying several
Figure 12.4
instruments. AP Photo/NASA.

were downward-pointing, so the camera and spectrometer were not usable

on the surface, but the GRS continued to operate and provided additional
data while NEAR was landed. The order of operations performed by NEAR,
with a rendezvous phase allowing detailed mapping and informing the
choice of a landing site, is seen as a model for most future small body lander


No matter how sophisticated the instruments carried on a mission, they are

inevitably less capable than the ones found in earthbound laboratories. This
is because the amount of preparation and lead time for missions is suffi-
ciently long that even the most modern equipment can become outdated by
the time of launch. In addition, there are special requirements for equip-
ment on spacecraft such that the most state-of-the-art instrumentation has
typically not been cleared for flight. As a result, there is a great impetus to
bring samples back from planetary bodies to Earth, where the most up-to-
date laboratory equipment can be used for analysis.
The first sample return missions were the Apollo landings on the Moon.
These returned nearly 400 kg of material from the Moon, and analysis of
Apollo samples continues to this day. Humans are not likely to be on any
small bodies in the coming decades, so any sample return would be robotic.
There have been robotic sample returns from the Moon, with the Soviet
Luna 16, Luna 20, and Luna 24 missions bringing back roughly 300 g total.
Sample returns have been proposed for Mars as well, perhaps occurring
sometime in the 2020s.
Because they can contain material that has been largely untouched since
the formation of the solar system, sample returns from asteroids and com-
ets are of particular interest to scientists. The large amount of meteoritic
material we have for study on Earth has been affected by the terrestrial envi-
ronment, and so is not as pristine as something collected from an asteroidal
surface. In addition, the collection of meteorites on Earth is biased; we
know, for instance, that very weak material will not be able to survive the
passage through the Earths atmosphere that meteorites experience. This is
true of the unconsolidated regolith of objects that produce meteorites as
well as some parent bodies that may not be represented at all. This is also
true of cometswe do not have any meteorites that scientists suspect are
from comets, and therefore any samples must be retrieved via spacecraft. As
a result, sample returns from the surfaces of a comet and an asteroid are
ranked as high priorities by NASA and have been the subject of intense in-
terest by European and Japanese scientists as well.
The Apollo and Luna sample returns both involved landers drilling into
the lunar surface. The small body sample returns that have flown to date
have used different approaches. The Hayabusa mission planners chose not
Spacecraft Missions  165

to land on Itokawa because of the landing difficulties mentioned previ-

ously. Instead, they used a touch and go strategy, where sample collec-
tion would occur very quickly, during a short interval when part of
Hayabusa was in contact with the asteroid surface. During that interval,
Hayabusa was designed to fire a projectile at Itokawas surface, knocking
off material and collecting it for the ride back to Earth. Due to problems
with the mission, it is still uncertain whether Hayabusa was able to collect
any samples, although this question will be resolved upon Hayabusas
return to Earth in 2010.
Due to the nature of comets, sample returns from these objects can be
both easier and harder than for asteroids. For instance, the fact that a comet
is constantly shedding grains of dust into space means that a landing is not
technically required in order to collect samples. NASAs Stardust spacecraft
took advantage of this in 2004 by flying through the coma of Comet Wild 2
carrying sample collectors made of aerogel. Aerogel is an exceedingly light-
weight material specially made with a very low density (0.002 g/cm3, over
1,000 times less dense than typical rocks). By using aerogel, scientists were
able to slow down the coma particles from over 6 km/s, the relative speed of
the comet and spacecraft, and capture the samples without destroying
them. Stardust returned roughly a million dust particles, the vast majority
much smaller than 100 mm in size. While analysis is ongoing and will con-
tinue for years, initial results suggest that Wild 2 contains grains that origi-
nated over a surprising range of distances in the solar system, with some
possibly originating from outside the solar system entirely.
A sample from a cometary nucleus is even more desirable than a coma
sample. The material found in the coma is only the silicate fraction of the
comets composition; the ices are absent after their sublimation. Better still,

Figure 12.5 The sample return mission Stardust used aerogel to slow down and capture
dust from Comet Wild 2. Shown here is a closeup of the Stardust aerogel, with a track
due to a captured dust grain visible just left of the center of the image. These grains are
extracted one at a time and analyzed, with the analysis still under way. NASA.

from a scientists point of view, would be a sample from beneath the surface of
a comet, which would contain the least-processed material. This, however,
would potentially run the risk of creating an outburst on the comet that could
put the mission in danger. Regardless, sample returns from a comet nucleus
have been identified as high priority by space programs around the world.


While humans wont be standing on the surface of any small bodies in the
next few decades, there have been proposals over the last 40 years to send
astronauts (or cosmonauts) to visit asteroids. The benefits and drawbacks
to human exploration are complicated, and there are strong feelings among
both opponents and proponents. It is generally agreed that the Apollo astro-
nauts were able to quickly identify interesting areas for exploration, and
their ability to be flexible and improvise allowed problems to quickly be
solved and on-site changes to be made. However, robotic missions are
much cheaper and simpler than those that carry crews, and the failure of a
robotic mission never has stakes or consequences as dire as the failure of a
mission carrying people.
A piloted mission to an asteroid has been advocated as a natural stage in
the exploration of Mars in particular and the solar system in general. This
advocacy centers on the idea that asteroid missions can be intermediate in
length between lunar missions and the flight time to Mars, allowing some
systems to be tested in conditions that are not as demanding as a Martian
mission. Over the long term, asteroids could be used for in-situ resource
utilization (ISRU), where astronauts process asteroidal regolith to release
water, oxygen, or metals that are useful for life support. In a similar vein,
Phobos, the inner satellite of Mars and likely a captured asteroid, has been
proposed as a base for the exploration of Mars.
A second rationale for human exploration of the small bodies has already
appeared in moviesas a way to prevent the impact of a near-Earth object
(NEO). In this case, some argue that the greater flexibility and problem-
solving that humans showed on the Moon would be more than worth the
extra expense, given the potential result of a large impact.


Although they are not missions to small bodies, orbiting satellite observatories
have also played a role in small body studies. Observing from space removes
the interference from the Earths atmosphere, whether from the risk of clouds
or bad weather, or the restricted wavelength range observable because of
absorptions due to water, methane, and carbon dioxide in our atmosphere, or
just the small-scale wind and turbulence that makes stars twinkle.
Spacecraft Missions  167

The most famous orbiting observatory is the Hubble Space Telescope

(HST), which has observed asteroids, comets, and dwarf planets since its
launch in the late 1980s. It has been used to discover satellites of Pluto, to
make multiwavelength maps of the asteroid Vesta, and to observe the
breakup and impact into Jupiter of Comet Shoemaker-Levy 9, among many
other observations. Other observatories that cover other wavelengths have
also played large roles in furthering small bodies research. The infrared
spectral region can be quite difficult to observe from Earth, but is an impor-
tant one for studying composition and can be used to determine the diame-
ters of objects. Two spaceborne infrared observatories, the Infrared
Astronomical Satellite (IRAS) and the Spitzer Space Telescope took very
large amounts of asteroid data, both as purposeful targets and serendipi-
tously (in much the same manner that astrophysicists in the early 1900s
found their photographs of galaxies littered with previously unknown aster-
oids). The IRAS and Spitzer data continue to be used to this day.
There have been proposals to launch future observatories specifically to
search for and catalog NEOs. With the ability to observe 24 hours/day and
in all directions, such a satellite could more quickly find potentially hazard-
ous objects than using only earthbound telescopes. Such proposals often
suggest using telescopes sensitive to the infrared region, where NEOs are
very bright relative to stars, rather than visible light. Studies have shown
that to be most effective, such a telescope would orbit the Sun at Venuss
distance so that objects interior to Earths orbit could be more easily
observed. While such observatories have not been approved, the first NEO-
observing satellite, a Canadian effort that is designed to orbit the Earth,
may launch early in the next decade.


There are obviously a large number of ways that spacecraft missions to

small bodies extend the capabilities of scientists and provide data that are
difficult or impossible to obtain from the Earth. However, there are also
drawbacks to spacecraft missions compared to other ways of studying the
asteroids, comets, and dwarf planets.
The first and perhaps most obvious is cost. The missions that the United
States has launched in the recent past to small bodies have had price tags of
roughly 400500 million U.S. dollars. The New Horizons mission to Pluto
has cost even more. Budgetary information for non-U.S. missions is often
not presented in the same fashion as U.S. missions, but building, launching,
and operating spacecraft costs many times more than building and operat-
ing even the largest telescopes.
In addition, space missions are exceedingly complex undertakings. Tele-
scopes and instruments on the Earth can be maintained relatively easily,
occasionally as simply as by having someone with a screwdriver do a

half-hour of work. Spacecraft, on the other hand, are required to function

for years, with no hope of physical intervention. They must be able to work
for long durations and designed to handle a wide range of conditions. Com-
ponents that work perfectly on their own can sometimes give unexpected
results when used together, and sometimes simple bad luck can intervene.
Because of the cost, there is pressure to ensure mission success, although
truly ensuring mission success would make costs skyrocket. NASA in partic-
ular has at times emphasized cost concerns over reliability, preferring a
steady stream of cheaper missions, some of which fail, to infrequent expen-
sive missions with greater likelihood of success. However, NASAs policies
have varied on this point through the years.
A final issue is time. From the time mission ideas are first considered
until the time a mission is launched is typically several years. From launch
until data are returned is often an additional few years. The New Frontiers
mission to Pluto will spend nine years en route to its destinationit spent
roughly twice that amount of time in development. Because of this, and the
desire for reliability, the instruments and components on spacecraft are of-
ten older designs and have capabilities that are not nearly as advanced as
what might be commonly available. For instance, the sizes of hard drives
and processor speeds in on-board computers are usually nowhere near the
capabilities found in relatively cheap personal computers. For example, the
computing ability of the Galileo spacecraft, active until 2003, was similar to
that of the first home video game systems of the 1970s. Again, this contrasts
with the ability of ground-based observatories to quickly use new equip-
ment and take advantage of new findings.


While the foundation for our understanding of asteroids, comets, and dwarf
planets comes from centuries of ground-based observations, spacecraft visits
have provided great advances in our knowledge using techniques impossible
to employ by any other means. Simple flybys provide close-up views of
objects impossible to obtain from Earth, and can allow masses to be meas-
ured as well as the composition of minerals on the surface. More complex
missions carry instruments that measure the relative abundances of elements
in a body and can precisely measure shapes as well as provide global imagery,
or even return pieces of the target to Earth for study in our laboratories.
With increased complexity come increased costs and an increased chance
of mission failure. These possibilities also arise with increased travel time
and distance from the Sun. While flybys of asteroids and comets have been
accomplished, and NEAR Shoemaker has landed on the asteroid Eros (and
an ESA cometary lander will operate in the next decade), the distance of the
dwarf planets means that only flybys are being considered at this time, with
the New Horizons mission to Pluto en route to a 2015 encounter.
Spacecraft Missions  169

As technology improves, difficult missions become easier and costs come

down. In coming years we look forward to sample returns from comets and
asteroids, flybys of additional objects, and further scientific breakthroughs
from space missions.


This is the Web site of Japans Hayabusa mission (most current-day missions
have a presence on the Internet):
This is the Web site of Europes Rosetta mission:
For those interested in taking part in the Stardust sample analysis, this Web site
provides that opportunity, as well as information about the Stardust mission:
These Web sites offer popular-level descriptions and explanations of the gamma-
ray and x-ray spectrometers. The MESSENGER mission to Mercury carries
these two instruments, which are commonly found on small bodies missions: and http://btc.
At this site, the user can access an online Hohmann Transfer calculator, with the
ability to input various starting and ending locations:
Missions of all kinds, from those that have flown to those that were more flights of
fancy, are described at these blogs, including small bodies missions as well
as those with other targets: and http://

In the preceding chapters, we have considered the asteroids, comets, and

dwarf planets from many different directions. We have looked at where they
are found, and what they are made of. We have thought about how they
formed and how they have evolved. In this final chapter, we will tackle the
question of how much they form a single coherent grouping, and to what
extent they are distinct from one another. We will also look at how they
relate to the other non-planetary bodies in the solar system.


The distinctions between asteroids and comets are in some ways clear and
in other ways quite blurry. As discussed in Chapter 1 and elsewhere, the
original, observational distinction is that asteroids have starlike appearan-
ces in a telescope (under typical conditions), while comets are fuzzy.
Additional research in the centuries since these definitions were created has
added a host of additional distinctions between typical comets and aster-
oids; as a general rule, asteroids are thought to be rocky, while comets have
significant amounts of ice.
During the time of the solar nebula, planetesimals formed with a variety
of compositions ranging from metal-rich and ice-free to mixtures of mate-
rials dominated by ice. The specific compositions were dependent upon
temperature, and followed a trend called the condensation sequence,
described in detail in Chapter 5. Looking at the distribution of spectral
types of asteroids (discussed in Chapter 7) with solar distance, we can


recognize a trend from an inner belt dominated by S-class asteroids through

a middle belt mostly populated by C-class asteroids, to P asteroids in the
outer belt. All these suggest a continuous trend in composition. The fact
that different groups are more common in different parts of the belt shows
that mixing between groups hasnt been a major factor, otherwise different
areas of the asteroid belt would have similar fractions of each spectral class.
If mixing hasnt been important, the obvious conclusion is that the different
groups formed close to where they can be found today. The D-class spectra
of Trojan asteroids seem to follow that trend, leading to the long-standing
interpretation that they formed near 5 AU, where they are today. Those
spectra of cometary nuclei that we currently have also have D-class spectra,
while Centaurs and TNOs have a mix of spectral slopes along with evidence
of water and methane ice on their surfaces.
However, with updated dynamical models over the past decade, astrono-
mers have realized that the asteroids and comets may have formed over a
variety of distances and moved from their formation locations. Indeed, the
Trojan asteroids are now thought to have formed perhaps as far from the
Sun as Uranus or Neptune and been perturbed into their current orbits.
While we do not think this is true of most main-belt asteroids, it is possible
that a few objects now in the asteroid belt might have formed elsewhere, or
that objects formed in places where there are no small bodies today might
have been luckily preserved far from their creation sites. As a result, we
might expect to find objects with characteristics we associate both with
comets and asteroids.
How could we recognize such objects? And how could we distinguish
between theories that predict the Trojan asteroids formed at 5 AU from
those that predict they formed at 1530 AU? The most straightforward
approach would be to obtain compositional information and compare it to
what we expect. Organic compounds are more stable farther from the Sun
than they are close to the Sun. When we compare the relative amounts of
carbon and other elements (like silicon or magnesium) that we see, we find
much more carbon in objects that formed in the outer solar system like
Comet Halley than we find in meteorites that formed in the inner solar sys-
tem. Although data are scarce, we might also expect the same to be true of
nitrogen, which is relatively abundant in the outer solar system as nitrogen
ice, ammonia (NH3), and cyanide compounds (those with a CN triple
bond). If objects from the outer solar system were scattered widely through
the solar system, we might expect that various comets, outer solar system
planetary satellites, and TNOs might have similar amounts of carbon and
nitrogen, or that there would be no consistent connection between compo-
sition and solar distance. Obtaining the data necessary to make firm con-
clusions will likely take several dedicated space missions, and may lie
decades in the future.
A study of bulk compositions at different solar distances could show
whether objects have moved far since their formation was completed. A
Interrelations  173

separate question is whether there was significant mixing very early on

and if objects are composed of only locally available material. It is recog-
nized that the asteroids and comets represent leftovers from planetary
formation, representing the conditions at locations in the inner and outer
solar system, respectively. In addition to the realization that objects
formed at 30 AU may now be found at 5 AU, and vice versa, results from
the Stardust mission to Comet Wild 2 suggest that at least in some cases
objects may be composed of mineral grains formed over a wide range of
distances. In fact, the silicate grains captured by Stardust are much more
similar to asteroidal material than what was expected from a comet. This
runs counter to established models of solar system formation that predict
the material that accretes to form bodies is collected from a relatively nar-
row range of solar distances.


While most large-scale shuffling of orbits occurred billions of years ago,

there are still individual objects today whose orbits are evolving, as touched
upon in Chapter 3. Jupiter has been the most significant actor in perturbing
and changing the orbits of small bodies throughout the history of the solar
system. Close passes to Jupiter can make objects from the outer solar system
into full-time residents of the inner solar system. However, by the laws of
orbital mechanics, some characteristics of the original orbit can still be
found in the new orbit.
As discussed in Chapter 3, an orbit can be described by a set of six orbital
elements. A complex combination of three of them, the semi-major axis (or
mean distance from the Sun), eccentricity (or how similar to a circle the
orbit is), and inclination (or the angle between the plane of the orbit and
the plane of the Earths orbit), along with the semi-major axis of Jupiter is
defined as the Tisserand parameter (Tj). The importance of the Tisserand
parameter is that it stays nearly constant before and after interactions with
Jupiter. For most comets, Tj is less than 3, while for most asteroids it is
greater than 3. A Trojan asteroid of Jupiter in a perfectly circular orbit co-
planar with Jupiter would have a Tj equal to 3. The Tisserand parameter is
often used as a means of interpreting whether or not an object began in the
outer solar system.
This is particularly useful for objects that share some characteristics with
different groups of small bodies. One such group is the Damocloids, named
after the object 5535 Damocles. Damocloids have orbits that look very
much like long-period comets, but show no evidence of cometary activity.
This has led to the interpretation that they are extinct comets, which once
had tails and comae but have exhausted their store of near-surface volatiles.
This would suggest that they are compositionally the same as comets even
though they might appear to be asteroids.

While the origin of Damocloids may be easy to understand, and their

orbits made an inner solar system origin obviously unlikely, the NEO popu-
lation hosts a more difficult situation to untangle. As described in earlier
chapters, bodies can evolve into near-Earth orbits from the asteroid belt or
the outer solar system. Because comets are expected to have icy interiors
compared to asteroids, and because a comet is expected to be weaker than
asteroids of the same size, different strategies might be employed to deflect
a threatening comet than a threatening asteroid. This makes the identifica-
tion of extinct comets in the NEO population of more than scientific inter-
est. It also adds to the desire to understand whether any objects with
compositions intermediate between asteroids and comets exist and whether
they can make their way into potentially hazardous orbits. The Tisserand
parameter is used in combination with reflectance spectra to try and iden-
tify the most likely cometary candidates.
One of the best extinct comet candidates in the NEO population is 3200
Phaethon. Its orbit is very similar to that of the Geminid meteor shower,
suggesting Phaethon is the Geminid parent body, where the material origi-
nates. All other known meteor shower parent bodies are comets, which has
led to the interpretation that Phaethon is likely very similar to comets.
However, other information about Phaethon points away from a cometary
interpretationit has never been seen to have any activity of any kind, and
its Tisserand parameter value is much higher than 3. Its reflectance spec-
trum is similar to the C-class asteroids of the mid-asteroid belt rather than
the steeply sloped spectra of outer-belt asteroids or cometary nuclei.
Another favored candidate is 4015 Wilson-Harrington. This NEO was dis-
covered in 1949 as a comet, and named Comet Wilson-Harrington after its
discoverers. It was lost shortly after, but its orbit was connected with that of
an asteroid discovered in 1979. It also has a C-class spectrum and a Tisser-
and parameter higher than 3. Wilson-Harrington has not shown any evi-
dence of cometary activity since its original discovery, leaving astronomers
puzzled about whether it is a comet that only occasionally has outbursts or
if it was the victim of a collision in 1949 that threw off dust and ejecta and
mimicked cometary activity.


Even within the main asteroid belt the distinction between asteroids and
comets can be blurred. Three objects have been found that orbit within the
asteroid belt but occasionally have comae and tails like comets. Studies of
their orbits show that these objects, called main-belt comets, must have been
formed at their current locations near 3 AU rather than being formed in the
outer solar system and transported to the inner solar system. For this reason,
some astronomers prefer the term activated asteroids to main-belt com-
ets since the latter could be interpreted as objects that started in the outer
Interrelations  175

solar system. As with Wilson-Harrington, the first-known main-belt

comet, 7968 Elst-Pizarro, was argued to possibly be the victim of a collision
(with dust thrown off from the impact spoofing a cometary coma) until a
second bout of cometary activity was seen several years after the first. The
discovery of two additional main-belt comets (2005 U1 and 1999 RE70)
made the collision hypothesis exceedingly unlikelygiven the estimated
impact rate in the asteroid belt, the odds that three objects on similar orbits
would suffer impacts within such a very short time are prohibitively long.
All three main-belt comets are members of the Themis dynamical family,
suggesting a common origin. Themis itself is much larger than the main-
belt comets, and has a spectrum like C-class asteroids. Themis itself shows
no sign of cometary activity. However, the current hypothesis is that all of
the members of the Themis family likely have ice at relatively shallow
depths, and that the ice became exposed on the main-belt comets. At this
point, the exposed ice began to sublime and carry off dust, just as with more
typical comets. Interestingly, reflectance spectra of Themis show some fea-
tures that have been interpreted as due to water ice. Observations are still
ongoing to try and determine how many of the more than 500 known The-
mis family objects show cometary activity. The existence of icy objects in
the Themis family also raises the possibility that other unrelated objects in
the outer asteroid belt could be ice-rich.



The satellites of Mars, Phobos and Deimos, were first imaged close-up by the
Mariner 9 spacecraft in the early 1970s. As small, irregularly shaped objects,
they were immediately suspected to be captured asteroids, and in the decades
before the Galileo spacecraft first encountered the asteroid Gaspra, knowledge
about these objects was held to be true of asteroids in general. For instance,
linear grooves are seen on Phobos and were interpreted as due to fractures
created during impacts. Similar features have also been seen on asteroids
like Gaspra and Eros (see Chapter 8), confirming this suspicion. Likewise,
some very large craters exist on Phobos and Deimos, much larger than were
once expected possibleStickney crater on Phobos is 9 km in diameter, while
Phobos itself is only 25 km across on average. There is evidence on Deimos
for a crater roughly 10 km across, while that body is only roughly 12 km
across on average. Here too, later observations of additional objects found
that asteroids have craters with diameters roughly the size of their own diam-
eters relatively commonly. However, in other ways Phobos and Deimos are
not like typical asteroids. Their location in orbit around Mars leads to re-
accumulation of ejecta that would escape if they were orbiting alone. This is
because while they have very low gravities and impact debris can relatively
easily escape Phobos and Deimos, the speed required to escape Marss gravity

is much higher. Typically the ejecta from impacts onto the satellites remains
in orbit around Mars, where it eventually re-impacts either Phobos or
Deimos. As a result, their regolith is much deeper than we would expect on
objects of the same size in the asteroid belt. Furthermore, there is some
evidence that ejecta from Deimos may be coating parts of Phobos, a type of
situation impossible for the majority of asteroids that are solo objects.
The origin of the Martian satellites is still not well understood. As detailed
dynamical models became available, it has been realized that it is exceedingly
difficult to create a scenario where Phobos and Deimos are captured objects,
either both at once or one at a time. Confusingly, the reflectance spectra of
Phobos and Deimos are most like outer-belt asteroids or even Trojan aster-
oids, which are among the most difficult objects for Mars to capture. The
dynamical problems facing capture scenarios have led to the suggestion that
the Martian satellites were themselves once part of Mars, blown off of its
surface in a giant impact similar to the one that created the Earths Moon.
Given the uncertainty surrounding the origin and composition of Pho-
bos and Deimos, their relationship to the asteroids and comets is as-yet
uncertain, even after several spacecraft have obtained data during opera-
tions around Mars. A Soviet attempt to land on Phobos during the 1980s
ended in failure, but a sample return from Phobos is planned by the Russian
Space Agency. It is hoped that in the coming decade we will better under-
stand the relationship between Phobos, Deimos, and other small bodies.


Pluto, Eris, and Ceres are the first three objects to have been classified as
dwarf planets. However, there are host of other objects that would also share
that definition if not for the fact that they are orbiting other planets and
thus disqualified. Indeed, some of the largest planetary satellites, like Gany-
mede and Titan, rival Mercury in size.
All the outer planets, as well as the Earth, have satellites in the diameter
range of roughly 1,0002,500 km. This size range also includes all of the
currently known dwarf planets. This creates an opportunity to study simi-
lar-sized objects with different compositions across a very wide range of so-
lar distances, from the rocky Moon and Io through Ceres rock/ice mix to
the icy satellites of Uranus and distant Eris. The study of comparative plan-
etology seeks to gain insight into solar system bodies by comparing them to
one another and pooling what we know about all of them rather than
studying each object independently. For instance, the similar sizes and for-
mation locations of Triton and Pluto have led many scientists to use knowl-
edge gained in the Voyager 2 visit to Triton and apply it to Pluto. The
computer models that are used to learn about the interiors of Ganymede
and Callisto have been used to study Ceres and large asteroids, in hopes of
determining how many of them may have icy mantles beneath their crusts.
Interrelations  177


As just discussed, the giant planets have a number of well-known, large sat-
ellites: Jupiter has Io, Europa, Ganymede and Callisto; Saturn has Titan and
Enceladus; and Neptune has Triton. However, each of these planets, as well
as Uranus, also has a large retinue of distant, small satellites. To contrast
them with the larger objects, and in view of their usually nonspherical
shapes, these are usually called irregular satellites.
Jupiter has nearly 50 irregular satellites outside the orbit of Callisto.
These seem to fall into at least three groups, seen in Figure 13.1. Perhaps the
most surprising characteristic of the outermost satellites is that they have
orbital inclinations greater than 90 degrees, which means that they orbit
Jupiter in a retrograde direction opposite from the majority of planetary
satellites in the solar system. The irregular satellites are extremely far from
Jupiter, the closest more than 5 times farther than Callisto (the outermost
of the large satellites) and most over 10 times farther, taking more than two
years to orbit Jupiter.
Like Phobos and Deimos, the irregular satellites of Jupiter are thought to
be captured objects rather than having formed in place. In particular, the
existence of the retrograde satellites seems to require capture, since all of
the studies of satellite formation show that objects formed in place must
revolve in a prograde direction. Dynamical studies show that it is much

Figure 13.1 The giant planets well over a hundred irregular satellites among them,
believed to be captured TNOs. The characteristics of those objects orbiting Jupiter are
shown here, with each point representing a satellite. With few exceptions they are small
(less than 20 km in diameter) and tens of millions of km from Jupiter, with some
taking years to orbit that planet. For comparison with the plot, Mercury is 0.4 AU from
the Sun. Illustration by Jeff Dixon.

easier for Jupiter to capture objects than Mars. The conditions that would
allow capture early in solar system history (such as an extended atmosphere
for Jupiter) no longer exist, although temporary captures are still happening
today; Comet Shoemaker-Levy 9 was a satellite of Jupiter for perhaps
20 years before its impact in 1994. The simplest explanation for the large
number of objects with similar orbits is that a few large objects were cap-
tured and then broke up either through impact or tidal stresses, resulting in
the swarm of satellites seen around Jupiter today.
We know relatively little about these objects other than their orbits. Most
of the irregular satellites are only a few kilometers in size, though a handful
approach 50 km and the largest, Himalia, is nearly 200 km in diameter.
Himalia was imaged by the Cassini and New Horizons spacecraft, but the
images were taken from a great distance and show little detail other than an
elongated shape. Detailed reflectance spectra are difficult to obtain because
of the small sizes and faintness of the Jovian irregulars, but there appears to
be at least two groupings, corresponding to the orbital groupings. The inner
Himalia group, which have prograde orbits also have spectra like C-class
asteroids. The outer retrograde group have spectra more like D-class aster-
oids and comets. There has been some suggestion that Himalia might have
hydrated minerals on its surface, again like the carbonaceous chondrites,
although that finding is yet to be confirmed by further study.

Figure 13.2 The best-observed irregular satellite is Phoebe. Cassini observed Phoebe in
2005, and scientists have concluded it is most likely a captured transneptunian object.
It is also conjectured that most irregular satellites had such origins. NASA/NSSDC.
Interrelations  179

The irregular satellites of Saturn similarly fall into several groups, based
on distance from Saturn and inclination. Again, these groups tend to share
similar colors, some like C-class asteroids, and some D-class. Again, this is
argued as evidence of an origin as captured TNOs. The spectral data has
been bolstered by the Cassini spacecrafts close pass of the irregular satellite
Phoebe. This pass obtained by far the best images of any irregular satellite
of Saturn (see Figure 13.2), and also obtained a measurement of Phoebes
density of roughly 1,600 kg/m3. This density suggests that Phoebe is roughly
half rock and half ice, much rockier than the larger and closer Saturnian
moons. This difference is further evidence that Phoebe did not form with
the inner satellites but was captured, and by extension the same is true of
the other irregular satellites.

Naming of Satellites
Until the mid-1800s, planetary satellites did not have a commonly accepted naming scheme. The
Galilean satellites of Jupiter (Io, Europa, Ganymede, and Callisto) were formally called JI, JII, JIII,
and JIV, their current names having been suggested soon after their discovery but never used in
practice. John Herschel, son of Uranuss discoverer, William Herschel, suggested names for Saturns
satellites drawn from Greek mythology, while Shakespeare provided the inspiration for the satel-
lites of Uranus.
Until the late 1900s, the number of planetary satellites was relatively small and few names were
needed (all were drawn from Western literature or myths). Only 65 planetary satellites were known
as the 1990s began (and 22 of those were discovered by the Voyager spacecraft), compared to the
166 known today. All these new discoveries were tiny irregular satellites.
After first trying to maintain the centuries-long precedent of naming satellites of Jupiter after
love interests of that Roman god, the International Astronomical Union (IAU) allowed moons to
be named for his descendents after discoveries outpaced his trysts. The irregular satellites of Saturn
have been named after giants and monsters, with a Norse-inspired group, a Gallic-inspired group,
and an Inuit-inspired group. Newly discovered Uranian satellites have continued to be named after
characters in Shakespeares work, while Neptunes moons are named for Greek sea deities.
Asteroidal and dwarf-planet satellites are named on a case-by-case basis, but usually the name
has some connection with the parent-body name (as opposed to the case for Uranus, for instance).
Occasionally, the satellite name has a whimsical aspect, as for instance with the 45 Eugenia system,
which has one satellite named Petit-Prince after the The Little Prince by Saint-Exupery and a more
recently detected satellite unofficially called Petit-Princess.

Uranus and Neptune also have sets of irregular satellites, though little is
known of most of them other than their orbits and estimates of their sizes.
The exception is Neptunes moon Triton, the seventh-largest satellite in the
entire solar system, with a diameter of 2,700 km (for comparison, the
Earths Moon is fifth largest, with a diameter of roughly 3,500 km). Triton
orbits Neptune in a retrograde orbit, a sign that it was a captured object.
Other satellites in the Neptune system, particularly Nereid, have orbits
that show evidence of being perturbed when Triton was captured. The

outermost satellites of Neptune have periods of more than 25 years, and

also have retrograde orbits. Indeed, it is possible that the capture of Triton
led to the loss of many satellites from Neptune, which alone among the
giant planets is missing a set of several regular, similarly sized satellites.


As mentioned previously, it is thought that 9099 percent of the mass of

TNOs and asteroids was ejected from the solar system. Therefore, in theory,
material from our solar system is currently traveling through interstellar
space and could find its way to other star systems. If the fraction of TNO
loss theorized for our solar system is also true for other solar systems, there
should be comets that formed around other stars occasionally making their
way into our solar system. Indeed, by one estimate there should be 1023
1024 such comets traveling through the galaxy (by comparison, there are
1011 stars in the galaxy).
Such interstellar comets should be easily detectable from their orbits.
Objects bound to the Sun have elliptical orbits. Objects with very large
semi-major axes but which come into the inner solar system, like long-
period comets, have large orbital eccentricities (see Chapter 3), and the
portions of the orbits we see look more like parabolas than ellipses. Objects
unbound to the Sun, however, will be approaching with enough speed to
escape from the Sun again (barring an exceptionally unlikely close pass to a
planet), and their orbits will follow hyperbolic paths. Despite centuries of
looking, no comet has ever been shown to have a hyperbolic orbit.

Figure 13.3 Orbits can have three shapes. Ellipses are the paths that all of the planets,
asteroids, TNOs, and dwarf planets follow. Circles are a kind of ellipse. Parabolas are
followed by long-period comets and indicate an object that is bound very weakly to the
Sun. Hyperbolic paths relative to the Sun have never been seen for a comet, and would
indicate an object visiting from outside the solar system. Illustration by Jeff Dixon.
Interrelations  181

Incredibly, however, astronomers are finding evidence of comets still or-

biting other stars. While any objects in the Kuiper belts of other star systems
are much too small to be seen from Earth, there is sufficient dust in those
Kuiper belts to allow their extents to be detected. Our own Kuiper belt is
thought to also have a similar dust ring, but ironically we are too close to
detect it because its light is too spread out. However, sharp-eyed readers
may have already seen the zodiacal light, a glow due to scattered light off of
asteroidal dust visible near the horizon near sunrise or sunset.
While visits from interstellar comets have never been identified, there is
very strong evidence for a stream of interstellar dust moving through our
solar system. This dust was identified by the Ulysses spacecraft whose main
mission is studying the Sun, with a secondary mission of studying inter-
planetary dust. However, it found some dust moving in a direction opposite
that expected for interplanetary dust, and with a speed of 26 km/s, faster
than the Suns escape velocity (1924 km/s over most of Ulysses eccentric
orbit) and interpreted as dust that originated outside of the solar system.
Beyond a distance of 3 AU, most of the dust with sizes of about 1 mm is
interstellar in origin rather than dust from our own solar system. The Star-
dust spacecraft sampled some of this interstellar dust and returned it to
Earth in 2006, with analysis still underway.


Comets and asteroids have long been seen as distinct objects, with comets
considered ice-rich objects from the outer solar system and asteroids rocky,
ice-poor objects from the inner solar system. However, it seems clear that
some objects have elements of both groups. Scattering of objects throughout
the solar system early in its history has potentially led to mixing, with objects
that formed at many different solar distances found in relatively close proxim-
ity today. The satellites of the major planets also share some characteristics
with the dwarf planets, asteroids, and comets, and in some cases they appear
to be captured asteroids or transneptunian objects. The small bodies also have
connections to other solar systemsscientists have observed evidence of
Kuiper belts around other stars, and dust from interstellar space is streaming
through our solar system and can be found in some primitive meteorites.


This Web site has a trove of information describing the irregular satellites of the
giant planets, put online by Scott Sheppard, who discovered many of these
The official NASA Web page for Phoebe, satellite of Saturn and possible captured

At this Web site, Henry Hsieh, who discovered the main-belt comets (or activated
asteroids), provides information about them, including links to more technical
Further information on the main-belt comets:
A list of Damocloids and technical papers about objects that straddle the line
between asteroids and comets is found at Dr. Yan Fernandez Web page: http://
A discussion of the Geminid meteor shower and Phaethon, its parent body, were
featured on the radio program Earth and Sky in 2007. This Web site has a tran-
script and audio file of the program:

Absolute magnitude. The magnitude (measure of brightness) of an object calcu-

lated for a distance of 1.0 astronomical units from the observer and the Sun. It is
related to the diameter of an object and often used as an estimate of size.
Absorption (light). One of the possible outcomes of an interaction between light
and matter. Absorbed light adds energy to the material with which it interacts.
Accretion. The growth of a body via collision, where the colliding objects remain
stuck together, resulting in an object with the combined mass of both.
Achondrite. A meteorite with no chondrules. The achondrites are one of the
major classes of meteorites, and include objects that have experienced high heat
and melting.
Activated asteroid. An object whose orbit is indistinguishable from a main-belt
asteroid but shows cometary activity. Such objects are also called main-belt
Aerogel. A very low-density material used to capture cometary and interplanetary
dust by the Stardust sample return mission.
Albedo. The fraction of light reflected from a surface, varying from 0 (perfectly
absorbing) to 1 (perfectly reflecting). Cometary nuclei and outer-belt asteroids
typically have albedos near 0.04, with asteroid albedos commonly ranging up to
Amor object. A near-Earth object (NEO) with an orbit that is always outside that
of the Earths.
Angle of repose. The greatest angle that a pile of loose material can support before
slumping. Cliffs greater than this angle are evidence that a material has some cohe-
sive strength.
Anhydrous mineral. A mineral lacking water in its structure.
Aphelion. The point in an objects orbit where it is furthest from the Sun.
Apoapse. The point in an objects orbit where it is furthest from its central body.
The aphelion is the same as the apoapse for objects that orbit the Sun.
Apohele object. A near-Earth object (NEO) with an orbit that is always inside that
of the Earths.
Apollo object. A near-Earth object (NEO) with an orbit that crosses the Earths
but is on average outside of the Earths.
Aqueous alteration. A type of metamorphism where chemical reactions in the
presence of water change the composition and type of minerals present, often to
include some minerals that incorporate water in their structure.


Areal mixture. A term describing a spectrum resulting from contributions from

distinct, unmixed regions. In an areal mixture, a given photon only interacts with a
single kind of material.
Astronomical unit (AU). The average distance from the Earth to the Sun, equal to
roughly 150 million km. The AU is very commonly used as a convenient measure
of distance in planetary studies.
Aten object. A near-Earth object (NEO) with an orbit that crosses the Earths but
is on average inside of the Earths.
Band center. The central wavelength of an absorption or emission band in a reflec-
tance or emission spectrum. The band center can be diagnostic of composition.
Band depth. The fraction of light absorbed at an absorption band center com-
pared to what would be expected if the band were absent. The band depth can be
diagnostic of the amount of a mineral that is present.
Blackbody radiation. Light given off by an object related to its temperature. A
perfect blackbody absorbs all light that falls on it, and does not exist in nature.
However, many objects act like blackbodies in some ways.
Blinking. A technique used to discover or characterize asteroids and comets, used
in the discovery of Pluto. Two photographs or images are aligned, with each visible
for a short period. Moving objects jump back and forth (or blink) between
frames, while stars and galaxies remain in place. Once done with rotating mirrors,
blinking is now commonly done with software.
Bodes law. A mathematical progression found in the eighteenth century that suc-
cessfully predicted the semi-major axes of Ceres and Uranus, but failed for Nep-
tune. Today Bodes law is considered by most astronomers to have little predictive
Bolide. A particularly bright meteor.
Calcium-aluminum-rich inclusion (CAI). Calcium and aluminum-rich materials
found in carbonaceous chondrite meteorites. These were the first solids formed in
solar system history, and their study has provided insight into the formation of the
solar system.
Carbonaceous chondrite. An important class of meteorites, often considered
among the best-preserved relics of solar system formation. Some carbonaceous
chondrites have been aqueously altered, and some include organic material.
Carbonates. A major mineral type on the Earth, containing CO3 as part of its
structure. Carbonates are also found in meteorites and have been found on Ceres.
Catastrophic collision. A collision destructive enough that the largest intact frag-
ment after the collision is less than half of the mass of the original target.
C-complex asteroid. One of the major taxonomic divisions of asteroid spectra. C
asteroids are the most common in the main belt as a whole and include Ceres,
Pallas, and Mathilde. They are often associated with carbonaceous chondrite
Centaur. A small body with an orbit between those of Jupiter and Neptune, often
crossing the orbit of one of the giant planets. Centaurs are thought to have origi-
nated in the transneptunian region and have compositions similar to comets.
Chondrite. A meteorite that contains chondrules. Chondrites are one of the major
classes of meteorites, and those meteorites most commonly seen to fall are chon-
drites. They have experienced relatively little heating and no melting since they
Glossary  185

Chondritic abundances. The relative amounts of elements found in chondrites,

particularly carbonaceous chondrites. Chondritic abundances are similar to solar
abundances when considering the elements found in rocks, although the Sun has
much more of some other elements like hydrogen and helium. It is believed that
the terrestrial planets have roughly chondritic abundances of the elements in bulk.
Chondrules. Small, glassy, round pieces of glass found in chondritic meteorites.
These are thought to have formed via a short-lived heating event very early in solar
system history.
Collisional families. See dynamical families.
Coma. The portion of a comet surrounding the nucleus, formed by and consisting
of gas and dust being lifted off of the nucleus. Comets close to the Sun typically
have their nuclei obscured by the coma, while at greater distances from the Sun the
coma may be absent.
Cometary jets. Regions on cometary nuclei where enhanced ice sublimation leads
to greater production of gas and dust in fan-like shapes, as well as the gas and dust
in those fans.
Condensation sequence. The order in which a hot gas with solar abundances
forms specific minerals as it cools down in ideal circumstances. The solar system is
thought to have roughly followed such a sequence.
Continuum. In spectroscopy, the level of a spectrum upon which no absorptions
or emissions are present. Band depths are measured relative to this continuum.
Crust. The outermost part of a differentiated object, usually formed by further
processing of the mantle. The crust, to be stable, has a lower average density than
the material below it.
Cryovulcanism. Low-temperature volcanic activity in which the magma is water.
Cryovulcanism is found on outer planet satellites and expected on larger transnep-
tunian objects.
Cybele group. A group of asteroids in a 3:2 mean-motion resonance with Jupiter.
Damocloid. An asteroidal object (that is, showing no cometary activity) found in
a cometary orbit.
Daughter molecule. A molecule formed in a cometary coma through the breakup
of other molecules through dissociation or other means.
D-class asteroid. One of the asteroid spectral classes, having steep spectral slopes
in the visible region. D-class spectra are associated with the outer asteroid belt as
well as cometary nuclei.
Delta-v. The amount of velocity change required for a spacecraft maneuver, or a
set of spacecraft maneuvers, often used to describe the total velocity change
required to get to a particular target.
D/H ratio. The ratio of deuterium (D) to hydrogen (H) found in a material.
Differentiation. The separation of an originally well-mixed object into layers, with
the densest material at the center. This usually occurs after heating and melting,
with separation into a core, mantle, and crust.
Dissociation. The separation of a molecule into constituent parts after interaction
with an energetic photon or particle.
Dust tail. The trail of dust from a comet, swept from the coma by solar radiation
forces. See also ion tail.
Dwarf planet. An object, according to the International Astronomical Union, that
is in hydrostatic equilibrium but is not large enough to have cleared its orbit.

Dynamical family. A group of objects that share similar orbits, suggesting that
they originated on the same object, broken up by collision.
Dynamical lifetime. The average amount of time an object will exist before it can
be expected to leave its current orbit.
Eccentricity. A measure of how circular an orbit is.
Ecliptic plane. The plane of the Earths orbit around the Sun.
Ejecta. Debris created in a collision and thrown from the impact site.
Emission. Light given off by an object, either because of its temperature or as
excited atoms return to their ground state.
Emission spectrum. The distribution of emitted light from an object, usually
showing regions of increased emission (or lines) on top of a continuum.
Encounter. In space missions, the period over which a target body is subject to
more-intense data collection.
Enstatite. The mineral name for magnesium-rich and iron-poor pyroxene. It is
one of the dominant minerals found in enstatite chondrite and aubrite meteorites.
Enstatite chondrite. One of the major groups of chondrite meteorites. The ensta-
tite chondrites are dominated by the mineral enstatite, and also contain iron-nickel
metal. They are associated with the M-class asteroids and thought to have formed
in the inner part of the asteroid belt.
Equilibrium condensation. A geochemical model of the early solar system that
assumes solids condensed out of a nebula with the same composition as the Sun,
with all chemical reactions going to completion.
Equilibrium saturation. The situation on a planetary, satellite, or small body sur-
face where the addition of new impact craters will erase some older craters, leaving
the overall number of craters on its surface roughly the same as before. Crater
counting can only reveal surface ages until this state is reached, after which only a
minimum age can be revealed.
Exobase. The level in an atmosphere above which atoms and molecules are more
likely to escape into space than to collide with other atmospheric particles.
Extinct comet. An object with a cometary history and overall cometary composi-
tion that once, but no longer, shows cometary activity.
Fall. A meteorite whose Earth entry was observed and was collected shortly
Find. A meteorite found without knowledge of its fall timing. Most finds occur in
Antarctica via concerted searching.
Flyby. An encounter where the spacecraft is never in orbit around the target. Usu-
ally flybys are relatively short in duration.
Fragment species. See daughter molecule.
Frost line. The distance from the Sun at which water ice becomes stable on an air-
less surface or as a free-floating crystal. This term is usually used in relation to the
early solar system rather than current times.
Gardening. Turnover and mixing of the top layers of regolith on an object due to
Gas tail. See ion tail.
Gravity assist. The use of a close planetary flyby to alter the path of a spacecraft.
Gravity assists help reduce the amount of fuel required to make large orbital
Glossary  187

Grooves. Linear structures seen on Phobos, Ida, and some other asteroidal objects.
While there are several interpretations for them, their exact formation mechanism
is not agreed upon.
Halley family comet. A comet with an orbital period less than 200 years and a Tis-
serand parameter less than 2.
HED meteorite. A group of achondrite meteorites including the howardite,
eucrite, and diogenite meteorite classes. The HED meteorites are thought to have
originated on Vesta.
Hilda Group. A group of outer-belt asteroids in a 3:2 mean-motion resonance
with Jupiter. These tend to be P- and D-class asteroids.
Hirayama family. See dynamical family.
Hungaria asteroids. Asteroids within 2.00 AU of the Sun, on the inner edge of the
asteroid belt. Hungarias are associated with E-class asteroids.
Hydrated minerals. Minerals that have water as part of their structure. Some also
include minerals with hydroxyl (OH) in their structure as hydrated minerals.
Hydrostatic equilibrium. The state of balance between the weight of overlying
material and the ability of deeper material to support it. Solar system objects in this
state are classified as planets or dwarf planets, depending upon additional dynami-
cal factors.
Hypervelocity impact. An impact with a speed higher than the speed of sound in
the target material. Most impacts in the solar system are hypervelocity impacts,
which have results very different from those at slower speeds.
Igneous. A rock or mineral type formed through the cooling of magma or lava.
Immature. A regolith that has recently formed and has not been affected by proc-
esses such as solar-wind implantation or micrometeorite bombardment.
Inclination. A measure of the angle between the plane containing an objects orbit
and the ecliptic plane.
Infrared (IR). A region of the electromagnetic spectrum stretching from roughly
750 nm (beyond the range of the human eye) to roughly 1 mm in wavelength.
In situ. Spacecraft investigations that occur on the surface of an object.
Interplanetary dust particle (IDP). Small grains derived from asteroidal colli-
sions or blown off of comets, and collected in the Earths atmosphere.
Intimate mixture. A mixture where different mineral components are in contact
with one another. In an intimate mixture, a given photon has interacted with more
than one kind of material.
Ion tail. A trail of ionized gas released from a comet, swept away by the solar wind.
Iron meteorite. One of the major groups of achondrite meteorites. Iron meteor-
ites formed either as the cores of differentiated bodies or in large impact-created
pools of molten rock.
Irregular satellite. One of the smaller, outer satellites of the giant planets. Most of
the irregular satellites are also irregularly shaped, and many have unusual orbits. It has
been speculated that they originated as captured asteroids or transneptunian objects.
Isothermal. Having the same temperature everywhere.
Isotopes. Atoms that have the same number of protons (thus belonging to the
same element) but a different number of neutrons (and thus different masses).
Jupiter family comet. Comets with a Tisserand parameter greater than 2. Jupiter
family comets have orbits that interact with Jupiter.

Kirkwood gaps. Regions in the asteroid belt that have been cleared of objects due
to mean-motion resonances with Jupiter.
K-T boundary. A layer in Earth rocks that separates material deposited in the Cre-
taceous Period from the Tertiary Period. It is believed a large impact occurred in
this time, with a mass extinction occurring as a result.
Kuiper belt object (KBO). A transneptunian object found in a low-inclination
orbit between 30 and roughly 55 AU.
Lag deposit. A surface layer on a comet that has become devoid of volatiles due to
their sublimation. A deep lag deposit can prevent further sublimation from deeper
Lagrangian points. Places in a two-body system where a third, small body can also
have a stable orbit.
Lander. A spacecraft that lands on an object, often conducting in situ studies.
LIDAR. Light Detection and Ranging, a technique used on spacecraft to measure
the distance to and shape of planets and small bodies.
Lightcurve. The change in brightness with time of an object. For asteroids this is
usually due to shape, though it could also include albedo effects.
Linear mixture. See areal mixture.
Lithophile. An element that prefers to be in rock rather than metal, and is left
behind in a rocky mantle when a metallic core is formed in an object.
Long-period comet. A comet with a period longer than 200 years.
Macroporosity. The amount of porosity in an object due to cracks and void space.
Magmatic iron. A type of iron meteorite formed during the differentiation of an
object, and once residing in a now-disrupted iron-nickel core.
Magnitude. A measure of the brightness of an object. Brighter objects have
numerically smaller magnitudes.
Main asteroid belt. A region between roughly 2 and 3.5 AU containing most aster-
oids. In general, orbits in the main belt are stable over billions of years.
Main-belt comet. See activated asteroid.
Mantle. A layer in a differentiated body that is less dense than the core. In the
inner solar system, the mantle is usually rocky over a metallic core, while in the
outer solar system it is typically icy over a rocky/metallic core.
Mars crossers. Asteroids whose orbits cross those of Mars but are not near-Earth
objects. These objects are not on stable orbits, and will become NEOs and be
removed from the solar system via impact on relatively short time spans.
Mass spectrometer. An instrument often carried on spacecraft missions that
measures the different atomic masses of particles, allowing elemental abundances
to be determined.
Mass wasting. The large-scale movement of material by gravity, such as landslides.
Mature. Regolith that has been subjected to high amounts of processes such as so-
lar wind implantation, ultraviolet light, and micrometeorite impacts. Mature rego-
lith can have properties quite different from immature regolith and rocks of the
same composition.
Mean motion resonance. An orbit with a period that is an exact fraction of the pe-
riod of a large planet. The smaller the numbers in a resonance, the stronger it is.
For instance, the 2:1 resonance (orbit period half that of a planet) is stronger than
the 13:7 resonance (orbit period 7/13 that of a planet). Kirkwood gaps are at loca-
tions of mean motion resonances.
Mesosiderite. An achondrite meteorite group containing both metals and silicates.
Glossary  189

Metamorphic. A rock type that has been subjected to heat, water, and/or pressure
sufficient to alter minerals and textures but not sufficient for melting.
Meteorites. A piece of an extraterrestrial object, usually an asteroid, that has
survived a fall through the Earths atmosphere and been recovered on the
Meteoroids. Objects that give rise to meteorites, as called before they enter the
Earths atmosphere. Often they are considered to be small (up to 50 m), otherwise
they are considered asteroids.
Meteors. The flash of light and trail associated with dust and small objects burning
up in the atmosphere.
Meteor shower. A period of high rates of meteors, seeming to originate from the
same point on the sky, caused when the Earths orbit intersects that of a dust
stream. Meteor showers recur on annual timescales.
Meteor storm. A meteor shower with an extremely high rate of meteor activity.
Miller-Urey experiment. An experiment in the 1950s that simulated the condi-
tions on the early Earth and demonstrated that organic materials could be created
from inorganic starting materials. This has also been argued as similar to the origin
of organic materials in meteorites.
Minerals. Collections of atoms in particular crystal structures.
Minor planets. A general term for asteroids and comets no longer in common use.
Mitigation. The prevention of asteroid or comet impacts into the Earth either by
diversion or destruction of the impactor.
Monolith. An asteroid or comet that is a single, unfractured piece.
Mutual event. A period when a satellite moves in front of or behind its primary
from the line of sight of an observer.
Near-Earth object (NEO). An object whose orbit comes within 0.3 AU of the
Earths orbit. NEOs are further subdivided into Aten, Amor, Apollo, and Apohele
depending on the specifics of their orbit.
Noble gas. One of several elements that do not form compounds, including
helium, neon, and argon.
Non-gravitational force. A force that alters a cometary or asteroidal orbit such
that it cannot be modeled as simply under the influence of gravity. Typical nongra-
vitational forces include forces due to the Yarkovsky effect, and due to gas and dust
loss on a comet.
Nonlinear mixture. See intimate mixture.
Non-magmatic iron. An iron meteorite that did not originate in the core of an
Nucleus. The central, solid portion of a comet. Nuclei are often obscured by the
cometary coma when close to the Sun.
Oblate. An object with a nearly spherical shape, but with a polar radius smaller
than the equatorial radius.
Obliquity. The angle between the equatorial plane of an object and the plane of its
Occultation. The blocking of one object by another. Asteroidal occultations, where
an asteroid moves in front of a star, have been used to determine shapes and sizes.
Olivine. A common mineral in terrestrial planets, small bodies, and meteorites,
containing iron, magnesium, silicon, and oxygen.
Oort cloud. A region up to 50,000 AU from the Sun hypothesized to be the source
of long-period comets.

Opposition. In astronomy, the point at which the Sun, Earth, and an object form
a straight line with the Earth between the Sun and object, when viewed from high
above the Earths north pole. It is also the point where the object is highest in the
sky at midnight when viewed from the Earth.
Optical depth. A measure of the amount of light scattered or absorbed by a coma
or atmosphere, which allows an estimate of the amount and kind of material
Orbit. The path taken by one body around another, dominated by the force of
Orbital elements. The set of six values needed to completely describe an orbit.
The standard (Keplerian) orbital elements are called semi-major axis, eccentric-
ity, inclination, mean anomaly, argument of periapse, and longitude of ascending
Orbital period. The amount of time required for an object to complete an orbit.
Ordinary chondrite. One of the major meteorite groups. Ordinary chondrites are
by far the most common falling meteorites. They have compositions of olivine, py-
roxene, and metal and remain relatively unaltered since they formed. They are asso-
ciated with the S-complex asteroids.
Organic material. Compounds containing both carbon and hydrogen, usually re-
stricted to solid material. It is found in carbonaceous chondrite meteorites and on
cometary surfaces, and is expected on some asteroidal and dwarf planet surfaces as
Oxides. Minerals containing oxygen but not silicon.
Palermo Scale. A quantitative measure of the hazard from a given NEO, taking
into account the time until a possible impact and the probability of impact. The
Palermo Scale is intended for specialist rather than public use.
Pallasite. One of a class of mesosiderite meteorites, known for having large crys-
tals of olivine mixed with metal.
Parallax. The change in relative position of objects caused by a change in viewing
geometry. Parallax has been used for centuries to determine the distance to objects
of interest.
Parent body. The object on which a meteorite originated, or the original object
from which a dynamical family was formed.
Parent molecule. A molecule that is broken up via dissociation or other means to
form daughter molecules.
Payload. The portion of a spacecraft that conducts scientific investigations.
Periapse. In an orbit, the closest point to the primary.
Perihelion. In an orbit, the closest point to the Sun.
Photometry. The study of brightness of astronomical objects.
Photons. Fundamental, massless particles of light.
Phyllosilicate. A class of minerals that are usually hydrous and formed via aque-
ous alteration.
Planetary embryos. Objects with diameters of roughly 1001,000 km that are
thought to have been common early in solar system history. Some planetary
embryos may still survive as dwarf planets.
Planetesimal. Objects with diameters of roughly 1 km that accreted early in solar
system history to form larger objects, including the planets.
Plutino. An object in a Pluto-like orbit, in the 2:3 mean-motion resonance with
Glossary  191

Plutoid. Defined by the International Astronomical Union (IAU) for objects that
are both dwarf planets and transneptunian objects.
Point source. An object that is too small to be seen as anything other than a point
of light in a telescope.
Ponds. Relatively flat, smooth, areas with fine-grained regolith on the surfaces of
Eros and Itokawa, and hypothesized to be present on other objects.
Porosity. The fraction of a rock, or an object, which is occupied by void
Potentially hazardous asteroid (PHA). An object with a size of at least 150 m that
passes within 0.05 AU of the Earths orbit.
Poynting-Robertson drag. A nongravitational force on dust grains that causes
them to spiral in toward the Sun.
Presolar grain. A dust grain that formed before the solar system formed. Presolar
grains have been found in some chondrites.
Pre-stellar Nebula. See protoplanetary nebula.
Primary. The most massive body in a system, around which other objects are con-
sidered to orbit. The Earth is the Moons primary, and the Sun is the Earths
Primitive achondrite. A group of meteorites that have experienced very limited
melting, placing them intermediate between the chondrites and achondrites in
many of their properties.
Production function. The distribution of impactor amounts and sizes for a
given object, and how that distribution changes with time. Knowledge of the
production function is required to allow surface ages to be calculated from
crater counts.
Prograde. Rotation or revolution in a counterclockwise direction when viewed
from above the north pole of an object or a system.
Protoplanetary nebula. A cloud of gas and dust from which planets and small
bodies form after a period of collapse and accretion.
Pyroxene. A common mineral found in terrestrial planets, small bodies, and mete-
orites. It contains silicon, oxygen, iron, magnesium, and calcium, in varying
Radiant. The point in the sky from which meteor showers appear to originate, and
from which they are named.
Radiation pressure. A nongravitational force on dust in the solar system, which
tends to sweep them away from the Sun.
Radiogenic heat. Heat generated from the decay of radioactive elements.
Radiometric ages. Ages calculated from the relative amounts of radioactive ele-
ments and their decay products.
Radionuclides. Radioactive nuclei of atoms.
Reflection (light). A change in the path of light at the surface of a material without
the light penetrating the material.
Refractory. A material that is stable at high temperatures and is expected to be
one of the first ones formed in solar system history.
Regolith. Loose, broken-up material found at the surface of an object. On airless
bodies, regolith is generated via collisions.
Regolith breccia. A meteorite type composed of regolith that has been hardened
into a rock. Regolith breccias give us much of the information we have about aster-
oidal surfaces.

Remote sensing. Data collection on an object using observations that are done
while not in contact with that object. Examples include imaging from orbit or
ground-based spectroscopy.
Rendezvous. A mission type in which the spacecraft expends delta-v to end up in
orbit around its target, or if its target is small and has weak gravity, in orbit around
the Sun in close proximity to its target.
Resolved. An object that is large enough to be detected as more than a point source.
Resonances. Regions in the solar system where orbits are not stable over long peri-
ods of time due to the periodic gravitational attraction of the planets. Objects in
resonances have their orbits modified by the planetary attraction, and usually end
up impacting one of the planets or the Sun, or being thrown entirely out of the so-
lar system.
Retrograde. Rotation or revolution in a clockwise direction when viewed from
above the north pole of an object or a system.
Rubble pile. An object consisting of pieces of many sizes that are in contact with
one another but are not bound by any forces other than gravity and friction. Rub-
ble piles have relatively low density compared to the material they are composed of.
Runaway growth. A period during the formation of the solar system when it is
thought that larger planetesimals grew at a much larger rate than smaller planetesi-
mals. This is hypothesized to have resulted in the formation of a relatively small
number of planetary embryos.
Sample return. A mission style where the spacecraft collects material from the tar-
get (either by landing or flying through a cloud of material) and returns it for anal-
ysis on Earth.
Scattered-disk object. A transneptunian object that has an eccentricity and/or in-
clination too high to place it in the Kuiper belt. It is thought that the scattered-disk
objects have had encounters with Neptune.
S-complex asteroid. One of the major spectral classes for asteroids. The S complex
includes most of the asteroids visited by spacecraft (Eros, Ida, Gaspra, Itokawa),
and most of the known NEOs. S asteroids dominate the inner asteroid belt and are
associated with ordinary chondrites and mesosiderites.
Secondary. In a binary or multiple system, those objects that are not the most
massive ones. The Moon is the Earths secondary. The Earth is one of a large num-
ber of secondaries orbiting the Sun.
Secular resonance. A region within the asteroid belt where the rate of precession
of certain orbital elements is synchronized with those of the giant planets. This
leads to rapid alteration of the asteroid orbits and removal of those objects from
the asteroid belt.
Sedimentary. A rock type composed of material deposited in layers before being
incorporated into the rock.
Seeing. A measure of the stillness of the atmosphere in terms of the angular resolu-
tion achievable with an arbitrarily sized telescope.
Semi-major axis. The mean orbital distance between an object and its primary.
Shattered. An object that has been thoroughly fractured, but whose pieces have
not subsequently moved relative to one another.
Short-period comet. A comet with an orbital period less than 200 years.
Siderophile. An element that prefers to be in metal rather than rock, and moves
into a metallic core when it is formed in an object.
Glossary  193

Silicates. A class of minerals that have silicon and oxygen. Silicates are the most
common minerals in the crust and mantle of the Earth, and are exceedingly com-
mon in meteorites and throughout the inner solar system and asteroid belt.
Size-frequency distribution. A description of how the amount of something
(usually craters or objects) changes depending upon their diameter.
Small solar system body (SSSB). Defined by the IAU as the set of solar system
objects that are neither planets nor dwarf planets.
Snow line. See frost line.
Solar abundances. The relative amounts of elements found in the Sun. Chondritic
abundances are similar to solar abundances when considering the elements found
in rocks, although the Sun has much more of some other elements like hydrogen
and helium.
Solar nebula. A region of gas and dust that according to current thought gave rise
to the Sun and planets after collapse and accretion.
Solid state greenhouse effect. A physical process in deep ice whereby light pene-
trates but the resulting heat cannot escape. This is thought to give rise to convec-
tion on some icy satellites and, in principle, TNOs, with surface features and
cryovulcanism possible results.
Space weathering. A general term for the processes that mature and alter regolith
properties, especially those that alter spectral properties as well.
Spectroscopy. The scientific field of studying the light from an object by separat-
ing it into its constituent wavelengths. Usually spectroscopy is used to determine
an objects composition.
Spectrum. The distribution of light at different wavelengths from an object. Also
sometimes more generally used to describe other distributions.
Sporadic meteor. A meteor not associated with a shower.
Stone. A meteorite that is neither stony-iron nor iron, including both chondrites
and achondrites.
Stony-iron. A meteorite group that includes pallasites and mesosiderites, with
both stony and metallic components.
Strength. The ability of a material to resist deformation or breakage.
Sublimation. The transitioning of a material (usually ice) from solid directly to
vapor form without passing through a liquid stage. This occurs at low pressures
and in a vacuum.
Synodic period. The time between oppositions of a body, or more generally, the time
between repetitions of the same geometry between a body, the Earth, and the Sun.
Tail. Gas and dust released from a cometary nucleus and swept out of the coma by
solar wind and radiation forces.
Terminal lunar cataclysm. A hypothesized period of increased collisions on the
Moon roughly 4 billion years ago.
Terminator. The dividing line on an object between sunlit areas and areas in
Thermal inertia. A quantitative measure of the ease or difficulty of heating and
cooling a particular substance. It is dependent upon several factors, including par-
ticle size and composition.
Thermal metamorphism. A type of metamorphism where chemical reactions
driven by heat change the composition and type of minerals present, although the
heat is not sufficient for melting.

Tisserand parameter. A measure of the influence of Jupiter upon an objects orbit,

which also allows some insight into previous orbits an object may have had. NEOs
with Tisserand parameter values less than 3 are considered possible comets.
Titius-Bode relation. See Bodes law.
Torino Scale. A qualitative measure of the hazard from a given NEO, based on its
size and the current knowledge of its orbit. The Torino Scale is intended for public
rather than scientific use.
Trajectory. The path followed by a spacecraft.
Transmitted light. In spectroscopy, light that has passed through a material.
Transneptunian objects (TNOs). Objects with orbits beyond that of Neptune.
Often the TNO group is considered to include the Neptune Trojan objects, but it is
not generally considered to include objects in the Oort cloud.
Trojan asteroid. An asteroid that orbits at L4 or L5 of a planet. Currently Jupiter,
Mars, and Neptune are all known to have Trojan objects.
T-Tauri stage. A phase early in the life of a star where powerful stellar winds
remove any remaining gas and dust from its nebula. This is generally considered
the end of star and planet formation.
Ultraviolet (UV). A region of the electromagnetic spectrum stretching from
roughly 400 nm (beyond the range of the human eye) to roughly 10 nm in
Undifferentiated. An object that has not differentiated. Undifferentiated material
has components with very different densities in close proximity to one another.
Unequilibrated. Material that contains minerals that formed under very different
temperature and pressure conditions, indicating that the material as a whole has
not come into equilibrium with one another.
Unresolved. See point source.
V-class asteroid. A member of an asteroid spectral class that has spectral proper-
ties very similar to the asteroid Vesta. V-class asteroids are associated with the HED
Visible light. Light with wavelengths between roughly 400 and 750 nm. Visible
light can be seen by the human eye, with differing colors ranging from violet (near
400 nm) through blue, green, orange, and yellow to red (near 750 nm).
Volatile. Material with a low boiling point, usually absent from inner solar system
bodies or present only in planetary atmospheres. Volatile material is the last to con-
dense in equilibrium condensation models.
Wavelength. The distance between successive crests or troughs in a wave. When
considering light, which has wave-like and non-wave-like properties, the wave-
length may be most easily considered to be a property related to energy.
X-complex asteroid. One of the major groupings in asteroid spectroscopy. The X
class includes asteroids with widely differing albedos and compositions. It is associ-
ated with aubrite meteorites, enstatite chondrites, iron meteorites, and some car-
bonaceous chondrites, among others.
Yarkovsky Effect. A nongravitational force related to asymmetric absorption and
reemission of light, altering an objects orbit. This force is only non-negligible for
rocky objects of roughly 10100 m in diameter, and even on those objects it can
take millions of years to make large-scale changes.
Zodiacal light. A glow in the pre-dawn or post-dusk sky due to sunlight scattering
off of the zodiacal cloud; dust generated from impacts in the asteroid belt and from
Annotated Bibliography


Beatty, J. Kelly, Carolyn Collins Petersen, and Andrew L. Chaikin, eds. The New
Solar System. 4th ed. Cambridge: Cambridge University Press, 1998.
An excellent collection of articles about planetary science topics, including
asteroids, comets, and transneptunian objects. It was published before the IAU
creation of the dwarf planet category.
Bell, Jim, and Jacqueline Mitton, eds. Asteroid Rendezvous: NEAR Shoemakers
Adventures at Eros. Cambridge: Cambridge University Press, 2002.
This volume includes contributions from the scientists involved in the mission
describing its conception and its findings.
Binzel, Richard P., series ed. University of Arizona Space Science Series. Tucson:
University of Arizona Press. 1979.
The University of Arizona Presss long-running Space Science Series publishes
cutting-edge research on planetary sciences. This series is relatively technical,
aimed at graduate school-level students and designed as a general reference for
professionals. The series currently has 30 volumes, but the books in the series
relevant to the topics discussed in this volume are listed below. The series is
periodically updated, so, for example, Asteroids III supersedes Asteroids II;
however, earlier versions of books can still provide interesting and useful infor-
mation and therefore are also included in the following list.
Asteroids III. Bottke, William F., Paolo Paolicchi, Richard P. Binzel, and Alberto
Cellino. 2002.
Asteroids II. Matthews, Mildred Shapley, Richard P. Binzel, and T. Gehrels.
Asteroids. Gehrels, T. 1979.
Comets II. Festou, Michel C., H. Uwe Keller, and Harold A. Weaver. 2004.
Comets. Wilkening, Laurel L., with the assistance of Mildred Shapley Mat-
thews. 1982.
Hazards Due to Comets and Asteroids. Gehrels, T. 1995.
Meteorites and the Early Solar System II. Lauretta, Dante S., and Harry Y.
McSween. 2006.
Meteorites and the Early Solar System. Kerridge, John F., and Mildred Shapley
Matthews. 1988.
Pluto and Charon. Stern, S. Alan, and David J. Tholen. 1998.


Protostars and Planets V. Reipurth, Bo, David Jewett, and Klaus Keil. 2007. This
volume is listed because it discusses solar system formation and includes chap-
ters touching on small bodies topics.
The Solar System Beyond Neptune. Barucci, M. A., H. Boehnhard, D. P. Cruik-
shank, and A. Morbidelli, with the assistance of Renee Dotson. 2008.
Consolmagno, Guy. Brother Astronomer: Adventures of a Vatican Scientist. New
York: McGraw-Hill, 2000.
Includes an account of Consolmagnos experiences on an Antarctic meteorite
expedition. The Web pages for the ANSMET expeditions can be found at
Davies, John. Beyond Pluto: Exploring the Outer Limits of the Solar System. Cam-
bridge: Cambridge University Press, 2001.
Written by an expert astronomer, this volume covers transneptunian objects
and centaurs in general.
Levy, David, and Stephen J. Edberg. Observing Comets, Asteroids, Meteors, and the
Zodiacal Light. Practical Astronomy Handbooks. Cambridge: Cambridge Univer-
sity Press, 1994.
Co-written by one of the most famous astronomers of recent times, David
Levy, this handbook was designed for those who wish to do observing them-
selves, with or without a telescope.
McSween, Harry Y. Meteorites and Their Parent Planets. 2nd ed. Cambridge: Cam-
bridge University Press, 1999.
McSween, a veteran meteorite researcher, wrote this useful overview of the
study of meteorites and what is learned from them.
Richardson, Derek, and Kevin Walsh. Binary Minor Planets. Annual Review of
Earth and Planetary Science. Palo Alto, CA: Annual Reviews, 2006.
An excellent overview of asteroidal satellites from a more technical angle.
Sagan, Carl, and Ann Druyan. Comet. New York: Random House, 1985.
The late Carl Sagan, one of the greatest popularizers of science in the twentieth
century, turns his attention to the small bodies.
Spencer, John, and Jacqueline Mitton, eds. The Great Comet Crash: The Collision of
Comet Shoemaker-Levy 9 and Jupiter. Cambridge: Cambridge University Press, 1995.
David Levys collaborators Carolyn and Gene Shoemaker contributed the fore-
word to this book about an object they discovered together, Comet Shoe-
maker-Levy 9, which later impacted Jupiter.
Stern, Alan, and Jacqueline Mitton. Pluto and Charon: Ice Worlds on the Ragged
Edge of the Solar System. New York: John Wiley and Sons, 1997.
Alan Stern is the principal investigator of the New Horizons mission to Pluto.
Warner, Brian, and Alan W. Harris. A Practical Guide to Lightcurve Photometry and
Analysis. New York: Springer-Verlag, 2006.
One of the most prolific lightcurve observers, Brian Warner, has coauthored
this in-depth and technical treatment of the subject.


Every year, dozens of articles about the newest research on asteroids, comets, trans-
neptunian objects, and dwarf planets are written in scientific journals. The journals
Annotated Bibliography  197

in which these articles are typically published include Icarus, Meteoritics and Plane-
tary Science, Astronomy and Astrophysics, and The Astronomical Journal, among
others. Lightcurve observations are often made by astronomy hobbyists, and their
work is often published in the Minor Planet Bulletin. Check recent issues of these
publications for articles on relevant topics. University and college libraries often have
subscriptions to the print or electronic versions of these journals, and some of them
also provide free access to older articles.


American Museum of Natural History:

This excellent set of online pages about asteroidal and planetary interiors and
differentiation is hosted by the American Museum of Natural History. A popu-
lar-level description of the interior of Ceres is produced at Astronomy maga-
zines Web site at
Antarctic Search for Meteorites (ANSMET) Program:
ANSMET is responsible for finding the vast majority of meteorites in the
worlds collections. Their homepage includes public information as well as
advice for those researchers chosen to join the hunt. The team occasionally has
been able to maintain Internet connections to the outside world and blog their
adventures. Blog entries from seasons up through at least 20082009 can be
found at
Asteroidal Satellite Discovery:
A description of the first ground-based detection of an asteroidal satellite by
the discovery team, along with movies and images.
A technical description of seismic shaking on asteroids, including a set of
explanatory images from NEAR Shoemaker and other missions.
The Basics of Light:
Contains additional detailed information on light.
Comets as Portents:
A recounting of how comets were sent as ill portents.
Cometary Studies:
A more general history of cometary studies.
A list of Damocloids and technical papers about objects that straddle the line
between asteroids and comets are found here at Dr. Yan Fernandez Web page.
Demotion of Ceres, Pallas, Juno, and Vesta from planet to asteroid: http://aa.usno.
Geminid Meteor Shower:
A discussion of the Geminid meteor shower and Phaethon, its parent body,
were featured on the radio program Earth and Sky in 2007. A transcript and
audio file can be found at this site.

History of the Status of Pluto:

Hohmann Transfer Calculator:
This online Hohmann Transfer calculator enables the user to input various
starting and ending locations.
Hubble Space Telescope:
A great deal of data from the Hubble Space Telescope has been used to study
protoplanetary disks around other stars. This Web site contains many of these
images as well as explanations for them in their gallery.
IAU: Video of IAU Debate:
IAU: Commentary on the IAU debate from a dissenting astronomer: http://
Irregular Satellites:
Atrove of information describing the irregular satellites of the giant planets
put online by Scott Sheppard, who has discovered many of these objects.
Main-Belt Comets:
The main-belt comets (or activated asteroids) were discovered by Henry
Hsieh, who provides more information, including links to more technical in-
formation, at this Web site. Further information on the MBCs can be found at
Dr. Anne Sprague maintains a set of information about the atmosphere of
Mercury at this Web site. Although it is not an asteroid, Mercurys atmosphere
is similar in character to what one might hypothetically expect on Vesta, if
conditions were right.
The MESSENGER mission to Mercury carries two instruments commonly
found on small bodies missions: a gamma-ray spectrometer and an x-ray spec-
trometer. Popular-level descriptions and explanations of these instruments are
found at these two Web sites.
Meteorites: University of Arizona:
Offers an abundance of information on meteorites.
Meteorites: Washington University:
More detailed material about meteorites.
Minor Planet Center:
Collects a large amount of data on near-earth objects, but is also designed to
help observers report and catalog possible new discoveries.
Minor Planet Center:
The Minor Planet Center is the central clearinghouse for collecting NEO discov-
eries and observations and calculating orbits. Information about these observa-
tions, including technical information about astronomers, is found at this site.
Mission Descriptions:
Missions of all kinds, from those that flew to those that were more flights of
fancy, are described at this blog, including small bodies missions as well as
those with other targets.
Annotated Bibliography  199

NASA: Ames Research Center

Offers additional NEO and impact information, including the Spaceguard
report that helped begin the current age of asteroid surveys.
NASA: Comet Shoemaker-Levy 9:
The breakup of Comet Shoemaker-Levy 9 led to a large amount of material
being posted on the then-new World Wide Web. This main page carries
updates from a wide variety of sources. A pre-impact summary of what was
expected from the Shoemaker-Levy impact can be found at http:// Other comet impacts, these into the
Sun, are described on,
along with images of them from the SOHO spacecraft. Another recent come-
tary breakup, that of Comet Schwassmann-Wachmann 3, was observed by the
Hubble Space Telescope, among many other observatories. The HST images
and descriptions can be seen at
NASA: Dawn Mission:
Has a set of links about Vesta and its interior.
NASA: Electromagnetic Spectrum:
A general introduction to the electromagnetic spectrum and its different
NASA: Genesis Mission:
Includes discussion of isotopic studies.
NASA: History of Asteroid Studies:
NASA: Jet Propulsion Laboratory:
The Jet Propulsion Laboratory (JPL) is in charge of NASAs effort to track and
characterize NEOs. This JPL site has a wide array of information, including
the text of a recent report to the United States Congress compiled by experts
and found at
NASA: JPL Radar Research:
The JPL radar research page contains a large amount of information on aster-
oid observations, including both more technical and more popular treatments,
and a large amount of images, movies, and links.
NASA: List of Planetary Missions:
This list of all NASA planetary missions offers links to NASA asteroid and
comet missions, on which sites additional official information can be found.
Most ongoing NASA missions have extensive education and public outreach
(E/PO) activities, which also have a web presence. The general NASA E/PO site
for planetary science
links to the specific types of objects described in this work.
NASA: Near-Earth Object Program:
Collects a large amount of news and general information on near-earth objects.
NASA: Page Defining Planets and Dwarf Planets:

NASA: Phoebe: page

The official NASA Web page for Phoebe, satellite of Saturn and possible cap-
tured TNO.
NASA: Positions and Orbits:
The positions and orbits of comets, asteroids, and dwarf planets are provided
by the NASA Jet Propulsion Laboratory. Views are available of the orbits of
individual objects, or positions of small bodies in large regions of the solar
NASA: Remote Sensing Data:
NASA maintains this central clearinghouse for a wide variety of remote sens-
ing data from both missions and Earth-based research.
NASA: Sedna Discovery:
The discovery of Sedna, thought to be the only Oort cloud object currently
known, is detailed at this site, with information about further Hubble Space
Telescope observations available at
Northern Arizona University:
This site provides further information about the minerals found in meteorites,
including chemical formulas and appearance.
Orbit Simulator:
This flash-based online orbit simulator includes several preset situations
(including a Trojan asteroid-like case) and the ability to input positions and
velocities to allow further exploration. A more-technical introduction to the
mathematics behind orbital calculations, including several further references,
is provided at
Planetary Society:
The Planetary Society offers this set of popular-level links on issues of near-
Earth objects and the impact hazard they pose.
Offers a more-detailed discussion of Plutos atmosphere, touching on other
TNO atmospheres and with sections focusing on small body satellites.
Scientific American article about the Planet Debate:
Sky & Telescope magazine:
A history of the scientific study of meteors.
Small Body Missions, non-U.S.:
MUSES_C.html and
Most current-day missions have a presence on the Internet. Two non-U.S.
small bodies missions with Web sites are Japans Hayabusa and Europes
Annotated Bibliography  201

Small Main Belt Asteroid Spectral Survey (SMASS):

An excellent resource for asteroid spectra.
Stardust Mission:
This site is for those interested in taking part in the Stardust sample analysis
and for obtaining information about the Stardust mission.
Tunguska Event:
A different look at the impact hazard is provided by scientist and artist Bill
Hartmann at this site, which details his attempt to make a painting of the
Tunguska Event as scientifically accurate as possible.
United States Geological Survey:
This USGS site has a detailed explanation of a wide variety of spectroscopic
techniques at a higher technical level than presented here. The USGS also pro-
vides a library of reflectance spectra for minerals at
University of Hawaii/David Jewett:
rubble.html. ( (http://www.ifa.hawaii.
The co-discoverer of the first Kuiper belt object, Dr. Dave Jewitt of the Univer-
sity of Hawaii, has this set of Web pages with information about comets and
dwarf planets. A simplified explanation of how lag deposits form on comets is
presented at the first site listed in this section. The latter two center on the
coma and tails of comets.
Windows to the Universe:
The Windows to the Universe Web site includes a set of Web pages about so-
lar system formation at various levels of detail (in Spanish as well as English),
and includes a reading list that offers books on solar system formation.

Absolute magnitude, 72 Charon, 3, 7, 73, 75, 79, 115, 133

Accretion, 6366 Chiron, 33, 138
Achondrites, 5051 Chondrites, 4849; carbonaceous, 4849, 51,
Albedo, 21, 30, 7174, 90, 95, 97, 110, 134, 8586, 95, 111, 178; ordinary, 49, 55, 90,
135, 152 97, 111, 162
Amors (NEO group), 39, 40 Chondritic abundances, 52, 54, 60
Apoheles (NEO group), 39, 40 Cladni, Ernst, 19
Apollo program, 53, 109, 164, 166 Classifications, 23
Apollos (NEO group), 39, 40 Clearing of neighborhoods, 8
Arecibo Observatory, xvii Cometary breakups, 118, 127, 136137
Aristotle, 14 Cometary interiors, 126127
Asteroid composition, 4751, 62, 8486, Comet Borrelly, 70, 113, 157, 158, 159.
9697 See also Deep Space 1
Asteroids: definition, 2, 910, 18, 171; Comet composition, 20, 62, 9192, 9899
discovery, 19, 20, 25, 144145, 167; Comet definition, 910, 171
possibility for atmospheres, 139140. Comet discovery, 16
See also individual objects Comet Halley, 14, 16, 24, 26, 33, 159, 172
Asteroid spectroscopy, 9596 Comets, 1, 910, 14, 16, 26, 3135, 40, 53,
Atens (NEO group), 39, 40 65, 81, 85, 112114, 118, 127, 148, 164,
165, 171174; atmospheres, 135137;
Binaries. See Satellites coma, 9, 33, 70, 89, 91, 98, 112, 135137,
Blackbody radiation, 90 157, 173; extinct, 113; interstellar, 180; jets,
Bodes law, 17 112; in NEO population, 40, 113, 174; as
omens, 14; tails, 89, 137138, 139, 174
Callisto, 176, 177 Comet Schwassmann-Wachmann, 3, 127,
Carbonaceous chondrites. See Chondrites 136137
Carbon dioxide, 62, 86, 91, 114, 127, 166 Comet Shoemaker-Levy, 9, 80, 127, 136, 144,
C-class asteroids, 95, 139, 172, 174, 175, 150, 167, 178
178, 179 Comet taxonomy, 9899
C-complex, 95. See also C-class asteroids Comet Tempel, 1, 113, 114, 158, 159. See
Centaurs, 33, 95, 138, 172 also Deep Impact
Ceres, xv, 4, 7, 8, 18, 48, 64, 70, 75, 95, 176; Comet Wild, 2, 109, 113, 165, 173. See also
atmosphere, 139; discovery, 1718; Stardust
interior, 125126; surface, 115116 Comparative planetology, 176
Charge-coupled devices (CCDs), 25 Condensation sequence, 6062, 136, 172, 176

204  INDEX

Craters and cratering, 4344, 102105, 113, Gravity tractor, 152

141, 142, 143, 175 Grooves, 107108
Cysat, Johann Baptist, 16
Halley Armada, xv, 159, 160
Dactyl, 76 Halley family comets, 33, 34, 35
Damocloids, 173 Haumea, 126, 134
D-class asteroids, 95, 98, 172, 178179 Hayabusa (mission), 26, 97, 162, 164165
Deep Impact (mission), 113, 159, 160 Herschel, William, 2, 16, 18
Deep Space 1, 157 Hirayama, Kirutsugu, 22, 36
Deimos. See Phobos and Deimos Hubble Space Telescope, 70, 127, 167
Density, 99, 120, 122123, 126, 134, 152, Hydrated minerals, 61, 178
161, 179 Hydrostatic equilibrium, 57
Differentiation, 50, 51, 85, 120, 122
Draper, Henry, 20 Ices, 8687, 94, 95, 112, 113, 127,
Dust, 31, 42, 101, 105, 174175; cometary, 135136
xvi, 30, 89, 112, 114, 135, 137, 165; Icy satellites, 175
extrasolar, 49, 58, 62, 84, 161; levitation, Ida, 6, 26, 75, 108, 158, 159. See also Galileo
105106; in solar system formation, 20, Impacts, xvixvii, 3637, 40, 4344, 53, 55,
61, 63, 65 79, 101103, 105, 118, 126, 134, 139,
Dust tail, 137 141143, 147148, 159; hazard, 147149;
Dynamical families, 22, 3638, 104, 126, mitigation, 150152
134, 139, 175. See also individual families Inclination, 27, 28, 32, 33, 34, 35, 38, 173,
Dysnomia, 75. See also Eris 177, 179
Infrared, 74, 87, 90, 94, 167
Earth, xvii, 13, 5, 6, 8, 43, 52, 53, 60, 61, 66, Infrared Astronomical Satellite (IRAS),
8385, 91, 93, 103, 105, 109, 114, 115, 119, 25, 167
120, 131, 132, 141143 International Astronomical Union (IAU),
Eccentricity (orbital element), 27, 28, 29, 34, 22; planetary definition, 59
64, 173 Interplanetary dust particles, 51, 101
Enceladus, 114, 177 Io, 103, 114, 176
Eris, 4, 62, 76, 99, 126, 145, 156 Iron, 48, 49, 52, 60, 84, 85, 99, 122, 126; in
Eros, 104, 105, 106, 107, 144, 162. See also space weathering, 94, 109
NEAR Shoemaker Iron meteorites, 19, 43, 47, 50, 85, 124, 148
European Space Agency (ESA), 158 Irregular satellites, 177180
Experiments, xvixvii Itokawa, 104, 105, 106, 162, 164165

Frost line, 6162 Japanese Space Agency (JAXA), 158

Jupiter, 1, 3, 6, 8, 13, 21, 26, 34, 35, 62, 63,
Galileo (mission), 26, 75, 127, 159, 168, 175 65, 79, 80, 127, 144, 173, 177, 178;
Galileo Galilei, 4, 15, 1617 influence on small body orbits, 19, 20,
Ganymede, 3, 176, 177 2829, 35, 64, 173
Gaspra, 26, 105, 108, 110, 158, 159, 175.
See also Galileo Keck Observatory, xvii
Gauss, Carl Friedrich, 18 Kepler, Johannes, 16, 27, 28
Gravity, 22, 23, 2829, 63, 64, 69, 74, 80, 81, Kirkwood, Daniel, 1920, 22, 35
102, 104, 106, 109, 117119, 131, 135, 140, Kirkwood gaps, 35, 36
152, 161, 162 Koronis family, 36, 37, 104
Gravity assist, 156 K-T Boundary, 142
Index  205

Kuiper, Gerard, 23 Near-Earth objects (NEOs), 21, 25, 3840,

Kuiper Belt, extrasolar, 181 144150, 167, 174; binaries, 7879, 80
Kuiper Belt object, 23, 3233, 35, 38, 64, 65, NEAR Shoemaker (mission), 97, 159, 162,
99, 118 163
Nebular hypothesis, 58
Lag deposit, 112113, 133, 139 Neptune, 3, 6, 8, 16, 22, 23, 28, 29, 33, 34,
Lagrange Points, 21, 34 40, 62, 63, 65, 66, 172, 179
Leonid meteor shower, 44, 45 New Horizons (mission), 115, 133, 167, 178
Le Verrier, Urbain, 23 Newton, Isaac, 16, 28
Lightcurve, 21, 75, 76, 78 Nitrogen, 52, 85, 86, 91, 115, 172; on Pluto,
Lowell, Percival, 23 132133; TNOs other than Pluto,
Lowell Observatory, 23 134135
Nongravitational forces, 2931, 81, 114
Magnetic fields, 128 Nucleus (comet), 30, 70, 137, 159, 165166
Main asteroid belt, 3536
Main belt comets, 138, 174175 Olbers, Heinrich William, 18, 19
Major planets, 14, 13, 17 Olivine, 50, 51, 60, 84, 92, 93, 94, 111, 122
Makemake, 134 Oort, Jan, 31
Mars, 1, 2, 6, 8, 13, 21, 22, 26, 31, 34, 35, 48, Oort Cloud, 3132, 33, 65, 99, 113, 136
85, 94, 103, 108, 118, 132, 156, 162, 163, Orbital elements, 17, 27, 28, 29, 32, 33, 34,
166, 176 35, 38, 64, 173, 177, 179
Mass wasting, 109 Ordinary chondrites. See Chondrites
Mathilde, 95, 103, 158, 159, 164. See also Organic material, 8586, 90, 91, 9899, 112,
NEAR Shoemaker 172
M-class asteroids, 95
Mercury, 1, 2, 3, 8, 13, 18, 21, 28, 40, 118, Pallas, 51, 64, 70, 126; discovery, 18
140, 176 Parallax, 1415
Messier, Charles, 16 Parent body processes, 5456
Meteorites, xvi, 19, 43, 4652, 141; Perseid meteor shower, 44, 45
classification, 4751; collection, 4647; Phaethon, 174
limits of study, 164 Phobos and Deimos, 95, 107108, 156, 167,
Meteors, 43, 4445 175176
Methane, 62, 86, 91, 98, 114, 134, 135, 166, Phoebe (satellite of Saturn), 178179
172; in Plutos atmosphere, 132133; Photography, xvi, 20, 21, 25, 144145
TNOs other than Pluto, 134135 Photometry, 21, 71
Minerals, 8387 Phyllosilicates, 85
Missions: flybys, 158160; landers, 162164; Piazzi, Giuseppi, 1718
limitations, 167168; rationale, 157; Planet-crossers, 33, 3839
rendezvous, 161162; sample return, Plutinos, 33, 35
164165; types, 157166 Pluto, 2, 34, 8, 23, 64, 70, 75; atmosphere,
Modeling, xvii 132133; discovery, 144; interior, 126;
Moon, 1, 2, 13, 15, 18, 20, 26, 37, 38, 47, 48, orbit, 29, 33; satellites, 7576, 79, 115, 167;
53, 65, 70, 79, 84, 95, 97, 102, 104106, surface, 114
109, 111, 140, 145, 155, 162, 164, 166, 176, Ponds, 106107
179 Porosity, 99, 118120, 123, 126
Potentially hazardous asteroids (PHA),
Naming objects, 4, 76 39, 40
NASA, 144, 152, 157, 164, 168 Poynting-Robertson drag, 31
206  INDEX

Presolar grains, 49, 51 Stardust (mission), 43, 51, 99, 109, 113,
Pyroxene, 50, 60, 84, 94, 97, 111 165, 173
Stony-iron meteorites, 47, 48, 5051
Radar, xvi, 31, 73, 75, 76, 99 Strength, 117120
Radiation pressure, 31, 137138 Sun, 1, 3, 8, 13, 30, 49, 51, 52, 55, 127; origin,
Radiogenic heating, 120122 58, 63, 65, 66
Radiometric age dating, 53 Synodic period, 145
Regolith, 55, 95, 102103, 104105, 106,
121, 164, 176 Telescopes, xvi, 70, 167
Resonances, 19, 2829, 33, 34, 35, 37, 38, Terminal lunar cataclysm, 65
40, 65 Themis family, 36, 37, 104, 139, 175
Rosetta (mission), 26, 108, 146, 159, Theory, xvii
162, 163 Thermal inertia, 30, 99
Rubble pile, 81, 108, 120 Tisserand Parameter, 34, 40, 173, 174
Russian Space Agency, 157, 176. See also Titan, 3, 85, 87, 176
Soviet Space Agency Tombaugh, Clyde, 23
Transneptunian objects (TNOs), 3, 10, 23,
Satellites, 7581; formation, 7981; 3233, 63, 92; atmospheres, 133135;
main belt, 7879; NEO, 78; TNO, 79. binaries, 75, 79, 81; interiors, 126;
See also Near-Earth object; Transneptu- taxonomy, 98
nian object Triton, 114, 115, 176, 179
Satellites of Pluto. See Pluto Trojan asteroid, 21, 34, 65, 172
Saturn, 1, 3, 13, 23, 28, 29, 34, 35, 65, 179 Tunguska event, 143
S-class asteroids, 95, 96, 172 Tycho Brahe, 14, 15, 21
S-complex, 95. See also S-class asteroids
Sedna, 32, 62, 134 Uranus, 2, 3, 8, 16, 17, 2223, 28, 34, 40, 65,
Semi-major axis, 17, 27, 28, 32, 33, 34, 35, 172, 176, 177, 179, 180
36, 38, 39, 66, 173, 180
Silicates, 49, 50, 54, 61, 62, 8485, 9495, Venus, 1, 2, 3, 6, 8, 13, 26, 28, 34, 60, 80, 103,
120, 165, 173 132, 163, 167
Size-frequency distribution, 36, 74, 104, 148; Vesta, 64, 70, 96, 97, 103, 139, 161, 162, 167;
as production function, 103 discovery, 18, 19, 48; HED meteorites, 50;
Sloan Digital Sky Survey (SDSS), 25 interior, 122125
Small bodies in popular culture, xi, 24, 150, Von Zach, Franz Xaver, 17, 18
166 Vulcanoids, 40
Small body shapes, 6975
Solar nebula, 58 Water ice, 30, 4647, 61, 75, 79, 8687, 108,
Soviet Space Agency, 25, 26, 157159, 112, 115116, 122, 123, 126, 127, 128, 133,
164, 176 136, 138, 139, 162, 175; ortho and para,
Spacecraft missions, xv, 2526 92; in space weathering, 95
Space weathering, 9495, 97, 109112 Wolf, Max, 20
Spectroscopy, 20, 8799; gas phase, 9092;
mixtures, 9394; solid state, 9294; X-complex, 95
taxonomy, 9596, 98, 99, 171172
Spitzer Space Telescope, 94, 136, 167 Yarkovsky Effect, 3031, 37, 38, 40, 152
About the Author

ANDREW S. RIVKIN was born in New York in 1969. As a boy, the Viking
and Voyager missions led to an interest in astronomy that lasts to this day. A
graduate of MIT with a doctorate in planetary sciences from the University
of Arizona, Rivkin now works at the Johns Hopkins University Applied
Physics Laboratory, specializing in observing asteroids and analyzing their
compositions. When not at the telescope or studying data, hes likely at a
baseball game or listening to the Beatles.